INFORMATION TO USERS
This was produced from a copy of a document sent to us for microfilming. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the material submitted.
The following explanation of techniques is provided to help you understand markings or notations which may appear on this reproduction.
1. The sign or “target” for pages apparently lacking from the document photographed is "Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you of complete continuity.
2. When an image on the film is obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because of movement during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that should not have been filmed, you will find a good image of the page in the adjacent frame. If copyrighted materials were deleted you will find a target note listing the pages in the adjacent frame.
3. When a map, drawing or chart, etc., is part of the material being photo graphed the photographer has followed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand corner of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete.
4. For any illustrations that cannot be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In all cases we have filmed the best available copy.
University Microfilms International 300 N. ZEEB RD., ANN ARBOR, Ml 48106 8121845
P o x c e, E l is e o R a m ir e z
FACTORS INFLUENCING AGRICULTURAL KNOWLEDGE ADOPTION LEVEL, AVERAGE RICE YIELD, PARTICIPATION LEVEL, AND PERCEPTION LEVEL OF SMALL-SCALE RICE FARMERS: A CASE STUDY OF SAMAHANG NA Y O N MEMBERS IN LEYTE, PHILIPPINES
The Ohio State University PH.D. 1981
University Microfilms International300 N. Zeeb Road, Ann Arbor, M I 48106 PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check Vmark .
1. Glossy photographs or pages______
2. Colored illustrations, paper or print______
3. Photographs with dark background______
4. Illustrations are poor copy______
5. Pages with black marks, not original copy______
6. Print shows through as there is text on both sides of page______
7. Indistinct, broken or small print on several pages\/
8. Print exceeds margin requirements______
9. Tightly bound copy with print lost in spine______
10. Computer printout pages with indistinct print______
11. Page(s)______lacking when material received, and not available from school or author.
12. Page(s)______seem to be missing in numbering only as text follows.
13. Two pages numbered______. Text follows.
14. Curling and wrinkled pages______
15. Other______
University Microfilms International FACTORS INFLUENCING AGRICULTURAL KNOWLEDGE ADOPTION LEVEL,
AVERAGE RICE YIELD, PARTICIPATION LEVEL, AND
PERCEPTION LEVEL OF SMALL-SCALE RICE FARMERS
A Case Study of Samahang Nayon Members in
Leyte, Philippines
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Eliseo R. Ponce
*****
The Ohio State University
1981
Reading Committee: Approved by
Dr. J. R. Warmbrod Dr. G. H. Phillips Dr. L. H. Newcomb Adviser department of Agricultural Education ACKNOWLEDGMENTS
The preparation of this dissertation has been one of the most rewarding experiences in my graduate study. First, it has allowed me to integrate and apply the research knowl edge I learned in my classes, giving me a sense of concrete accomplishment. More important, however, it has given me the opportunity to understand better the meaning of human good ness. A number of persons went out of their way to help make this dissertation possible.
The important assistance received from my colleagues and friends at the Visayas State College of Agriculture in
Leyte, Philippines is gratefully acknowledged. Manuel
Ancheta, Sarah Ancheta, Salvador Dagoy, Fe Jazon, Edith
Ventula, and Alberto Ricarte graciously donated countless hours beginning with the pretesting of the instrument until the data collection stage. Additionally, the President of the College, F. A. Bernardo, was most kind in helping obtain partial support for this research from EDPITAF.
Like most quantitative social research, the computer was extensively used in this study. Special credit is due to
Pattie Costello and Fred Reuland of the Statistical Labora tory in the Department of Statistics for their help in the use of special computer programs, including the interpreta tion of the "tons" of computer printouts. Fred had been
ii most generous with his time in spite of his busy schedule as head of the laboratory.
Sincere appreciation is also extended to John Kennedy, professor of the Department of Educational Foundations and
Research, and to Edward Carter, visiting statistics profes
sor from the University of Guelph, for the help and guidance
in the use of multivariate statistical procedures in this
study.
Of course, this study would not have been possible without the sustained guidance and inspiration of my adviser,
J. R. Warmbrod, throughout my entire graduate study in this university. As a recognized outstanding professor of social research, he has been most influential in my decision to
specialize in quantitative methods.
The valued help of the other members of my Dissertation
Committee is also appreciated. L. H. Newcomb and G. H. Phil
lips had been most constructive in their comments and
suggestions.
Special thanks also goes to the Graduate School of The
Ohio State University for the Graduate Alumni Research Award and to my former adviser, R. McCormick, for his valued guid ance during the first year of my graduate study.
Finally, to my wife, Lucylen, and to my two daughters,
Lezyl and Lueli, a big "thank you" for the joy of discover ing that graduate study, in spite of certain difficulties, can be fun and enjoyable because they are there to under
stand, inspire, help, and learn with me. VITA
June 14, 1943 ...... Born-Malabuyoc, Cebu, Philippines.
1964 ...... B.S. in Agricultural Education Visayas State College of Agricul ture, Baybay, Leyte, Philippines.
1964-1968 ...... Secondary School Teacher Naval School of Fisheries Naval, Leyte, Philippines.
1968-1969 ...... Grantee, Shell Oil Scholarship, Philippines.
1969-1970 ...... Grantee, National Science Devel opment Board Scholarship, Philippines.
197 0 ...... "...... Master of Arts in Teaching University of the Philippines Diliman, Quezon City, Philippines.
197 0-1973 ...... Secondary School Teacher and Head, Vocational Education Depart ment, Cebu Agricultural College, Cebu City, Cebu, Philippines.
1973-1975 ...... Principal, Capiz Agricultural and Fishery School, Pontevedra, Capiz, Philippines.
1975-1977 ...... Assistant Professor and Coordina tor, Community Extension Service, Visayas State College of Agricul ture, Baybay, Leyte, Philippines.
1977-1978 ...... Grantee, World Bank Scholarship.
1978-1981 ...... Graduate Research Associate National Center for Research in Vocational Education, The Ohio State University, Columbus, Ohio, U.S.A.
iv FIELDS OF STUDY
Major Field: Agricultural Extension
Studies in Social Research Dr. J. Robert Warmbrod and Statistics Dr. John J. Kennedy Dr. Kent Schwirian
Studies in Rural Dr. G. Howard Phillips Development Dr. William Flinn Dr. Ted Napier
Studies in Extension Dr. David Jenkins Methods and Program Dr. Ralph Bender Development TABLE OF"CONTENTS
Page
ACKNOWLEDGMENTS ...... ii
VITA ...... iv
LIST OF TA B L E S ...... viii
LIST OF FIGURES ...... X
CHAPTER
I. INTRODUCTION ...... 1
The Philippine Situation ...... 6 Statement of the Problem and Objective of the S t u d y ...... 10 Delimitation and Scope of the Study . . . 20 Organization of the Dissertation ...... 20
II. CONCEPTUAL FRAMEWORK ...... 21
The Consensus Perspective ...... 21 The Conflict Perspective ...... 28 C o n c l u s i o n ...... 36
III. RESEARCH METHODOLOGY ...... 38
Population and Sample ...... 38 Instrument Development ...... 40 Data Collection Procedure...... 45 Data Preparation and A n a l y s i s ...... 49
IV. FINDINGS AND DISCUSSION ...... 55
The Respondents: Two P r o f i l e s ...... 55 Relationships Between the Sets of Predictor Variables and the Set of Criterion Variables...... 79
vi TABLE OF CONTENTS (continued)
Page
CHAPTER
V. SUMMARY, CONCLUSIONS, AND IMPLICATIONS . . . 117
S u m m a r y ...... 117 Conclusions and Implications ...... 129
REFERENCES ...... 134
APPENDIX
A. MEANS AND LEAST SQUARE MEANS ...... 143
B. INTERVIEW SCHEDULE ...... 148
vii LIST OF TABLES
TABLES Page
1 Distribution of Respondents by Demographic Variables and Geographic Location ...... 57
2 Distribution of Respondents by Technical Factors and Geographic Location ...... 61
3 Distribution of Respondents by Political Factors and Geographic Location ...... 65
4 Distribution of Respondents by Physical Factors and Geographic Location ...... 66
5 Distribution of Respondents by Criterion Variables and Geographic Location ...... 69
6 Multivariate Analysis of Variance of All Numeric Variables by Geographic Location . 76
7 Summary of Significant Multiple Comparisons 77
8 Multivariate Analysis of Variance of All Criterion Variables by Geographic Location, Land Ownership, and Credit Source ...... 81
9 Summary of One-Way Analyses of Variance of Each Criterion Variable by Geographic Location, Land Ownership, and Credit Source 82
10 Summary of the Canonical Corelation Analyses Between the Set of Twenty-two Predictor Variables and the Set of Four Criterion Variables ...... 93
11 Structure Coefficients from the Canonical Analysis of the Set of Twenty-two Numeric Predictor Variables and the Set of Four Criterion Variables (Eastern Leyte) .... 95
12 Structure Coefficients from the Canonical Analyses of Each Subset of the Numeric Predictor Variables and the Set of Four Criterion Variables (Eastern Leyte) .... 101
viii TABLES Page
13 Structure Coefficients from the Canonical Analysis of the Set of Twenty-two Numeric Predictor Variables and the Set of Four Criterion Variables (Western Leyte) .... 105
14 Structure Coefficients from the Canonical Analyses of Each Subset of the Numeric Predictor Variables and the Set of Four Criterion Variables (Western Leyte) .... Ill
15 A Summary Profile of Eastern and Western Leyte Farmers ...... 121
16 Comparative Effects of the Twenty-two Numeric Predictor Variables on the Set of Four Criterion Variables by Geographic L o c a t i o n ...... 130
ix LIST OF FIGURES
Page
Figure
1 Map of the Republic of the Philippines .... 7
2 Graphic presentation of the consensus strategy for small-scale farmer devel opment ...... 26
3 Comparative graphic presentation of the consensus and conflict perspectives of s o c i e t y ...... 30
4 Graphic presentation of the conflict strategy for small-scale farmer devel opment ...... 35
5 The interaction effects of land ownership and credit source on average rice yield, participation level, and perception level . . 84
6 The interaction effects of land ownership by credit source on average rice yield, participation level, and perception level . . 85
7 The interaction effect of land ownership and geographic location on perception level. . 88
8 Graphic presentation of the multivariate relationships between the set of predictor variables and the set of criterion varia bles on the resultant structure coefficients from the reduced canonical analyses on eastern Leyte ...... 103
9 Graphic presentation of the multivariate relationships between the set of predictor variables and the set of criterion varia bles on the resultant structure coefficients from the reduced canonical analyses on western Leyte ...... 114
10 Matrix showing the comparative effects of the different subsets of predictor variables on the criterion variable set by geographic location ...... 116
x CHAPTER I
INTRODUCTION
One of the most serious problems confronting the world today is poverty— a condition which has continued to to per sist in both urban and rural areas in both industrial and developing countries. However, nowhere is it more vivid and serious as it is in the rural areas of low-income nations.
The World Bank estimates that 7 00 to 800 million people, 33 percent of the population of the Third World, are "economi cally deprived rural people" who seem to live in a perpetual cycle of poverty— a condition which, among other indicators, is characterized by low productivity, malnutrition, igno rance, and low life expectancy (Biggs, 1974, p. 8; Woods,
1975, p. 147). Included in this category are "subsistence farmers who provide the food needs of roughly 50 percent of the world's entire population on 40 percent of the world's agricultural land" (Biggs, 1974, p. 8). These subsistence farmers operate "very small farms— small patches of land that hardly qualify as more than backyard gardens" in west ern countries like Australia, Canada, and the United States
(Owens, 1974, p. 21). It is estimated that the world's small farm families whose land holdings are less than five hectares
(twelve acres) represent over one billion individuals (Woods,
1975, p. 147). The sheer magnitude of the population of small farmers makes them a logical component in national development pro grams among the developing nations. However, this is not so.
Past government programs have largely ignored this segment of the rural population (Biggs, 1974; Woods, 1975). In fact, it was only during the mid 1970's that small farmers were considered critical factors in rural development. Woods
(1975) observed:
A new topic is being discussed by national leaders and in national presses of the less developed countries of the world. The topic is the small farmer. The concern expressed for his well-being, for in the past it has been easy and convenient to ignore him. This new awareness is perhaps due to the massive numbers of small farmers and the impact of their problems on the rest of society. Perhaps it is the growing recognition that national growth is better measured by concerns of equity in distribution of income, a more equal sharing in the amenities of life, than by judgment relating to percentage increases in the gross national product (p. 147).
A brief review of the development philosophy and events during the last decade is presented for better understanding of this "new" development logic that has been expounded by the United Nations (Food and Agriculture Organization, 1975) and the World Bank (1974).
Development has been traditionally viewed as "sustained growth per capita income" (Oshima, 1967, p. 7). Agriculture development focused mainly on stimulating higher levels of production through diffusion of new technology. It is assumed that increasing agricultural productivity speeds the process of capital accumulation and, consequently, the extraction of surplus value for industrialization. In turn, the resulting increased industrialization absorbs the excess labor force from the countryside and generates additional demand for consumer goods, thus stimulating further agricultural deve lopment. The end product of which is the elimination of po verty and the reduction of economic inequality.
The farmers who have been the primary recipient of this technology-transfer-oriented development strategy are those who have superior education and resources— the progressive farmers. It is argued that, compared to the small farmers who have very little education and resources, the progres sive farmers are in the best position to demonstrate the immediate profitability of an innovation; thus, their farms often serve as local demonstration farms. They become the prime target of extension classes, soft credit, and other government assistance for increasing agricultural producti vity. It is anticipated that the benefits received by these immediate recipients will diffuse or trickle down to the less progressive sectors of the community. It will then create a ripple-effeet, thus accelerating not only agricultural deve lopment but also rural modernization.
Experiences among the developing countries, however, showed otherwise. While agricultural output has grown, rural poverty still persists and income inequality has, in fact, worsened. The relatively high economic growth rates obtained during the last decade have brought little benefit to *-he poorest segments of society. For example, in Mexico the ratio of income controlled by the top 20 percent compared to the bottom 20 was 10:1 in 1950 and 12:1 in 1969; in Brazil the ratio was 22:1 in 1960 and 25:1 in 1970 (Biggs, 1974, p. 5).
Productivity of small farmers has remained low in spite of demonstrated profitability of new agricultural innovations among progressive farmers. The trickle down theory of deve lopment "is proving utterly inadequate to the needs of the poorer halves of populations in developing countries" (Biggs,
1974, p. 2). The World Bank (1974) summed up the situation:
Emphasis on the development of the modern economic sector, providing employment to a small intensively trained elite leads to the neglect of the 60 to 80 percent of the population characterized by low productivity (p. 2) .
The growing social unrest in many developing nations in spite of economic growth, as shown by Latin America and
Asia, demonstrates one important fact: increased per capita
income alone does not guarantee the existence of a socially and politically stable nation. This insight has "helped turn the attention to development strategies which are directed
to sharing the benefits of growth as well as growth itself"
(World Bank, 1974, p. 14). This means that development pro grams aimed at the reduction of social and income inequali
ties should require and guarantee the full and active par
ticipation of all segments of society including equal access
to key social institutions and government services. Biggs (1974), a leading American proponent of strong small-farmer development programs, summed up the economic logic of this
"new" perspective:
There is increasing empirical evidence from both the developed and developing countries which cast considerable doubt upon the sig nificance of size economies in agriculture. For some years, there has been ample evidence attesting to the increased relationships between farm size and capital productivity. Since land and capital are the relatively scarce factors of production in developing countries, productive techniques and farm sizes which serve to increase the producti vities of these factors are consistent with economic efficiency. These results tend to refute the myth that small operating units are inefficient in the use of scarce re- * sources . . . the higher productivity levels will permit a portion of the extra output to be marketed through commercial channels. Not only will this add to the family's total income stream but also will assist in in tegrating the marginal families into the modern society and economy, an important aspect of the modernization process . . . A more egalitarian distribution of income will widen markets for consumer goods there by increasing the effective demand for industrial investment and encouraging in dustrial employment (p. 14).
It is against this background that world leaders recog nized the social and economic rationality of the foregoing development perspective; thus, during the United Nations
World Food Conference in Rome (Food and Agriculture Organi zation, 1975) a major resolution was passed which affirmed that there should be appropriate emphasis on:
A progressive social and structural transforma tion in agriculture . . . [and] an active parti cipation of the rural population, particularly small farmers and landless workers in the deve lopment process (p. 133). The Philippine Situation
The Republic of the Philippines is a country of 46 mil
lion people, situated 500 miles from Southeast Asia (Fig
ure 1). The country is composed of 7,107 islands with an
area of 115,800 square miles. Its people are predominantly
Malayan, "with sprinkling of Chinese, Indian, Arabic, and
Caucasian blood" (Christian Science Monitor, September 19,
1980, p, B2). Except perhaps for these features, the Philip
pines shares many of the social, economic, and political
problems confronting developing countries. Consider the
following statistics:
1. It has one of the highest birth rates in the world
with 2.6 percent annual rate of increase (Time,
March 2, 1981, p. 6).
2. The distribution of income is very unequal with the
lowest 20 percent of the families receiving less
than 5 percent of the total income and their share
declined over time from 4.5 percent in 1957 to 3.8
percent in 1971 (Institute of Economic Development
Research, 1976) .
3. A majority of the families (70 percent) reside in
the rural areas. Per capita income for rural fami
lies is US $287; urban families, US $618 (Szal,
1979, p. 90) .
4. Prices have risen sharply during the last few years
In the sixties inflation was 5 percent per year. In WvM.VAtijl.J.i
*>/a . Kindorfi *siavtA.
Figure 1. Map of the Republic of the Philippines. 1980, inflation hovers at 20 percent (Christian
Science Monitor, September 19, 1980, p. B2) .
The majority of the households are farm households situated in villages belonging to municipalities of less than 50,000 inhabitants and consisting of nuclear families occupying a single dwelling unit (Concepcion, 1972, p. 45).
In general, dwelling units (made of wood or bamboo with galvanized iron roofing or thatched nipa shingles) consist of two to three rooms, no piped-in water, with a kerosene lamp for lighting, and wood for cooking fuel. There are no food storage facilities such as refrigerators or ice boxes.
Most heads of rural households have completed primary or elementary education and are either farm owners with an average farm area of approximately three hectares (seven acres), tenants, or laborers (Castillo, 1975, p.264). With six as an average number of children, each person has a unit allotment of only one-third hectare (Concepcion, 1972, p. 44).
Castillo (1977), a leading Philippine rural sociolo gist, summed up the situation:
The Philippines, despite [its] Miss Universe . . .Fifth Avenue buildings, . . . plush hotels, and condominiums in Makati, is es sentially a nation of villages . . . with 70 percent of its population inhabiting rural and agricultural . . . communities . . . 60 percent of the families are low income and more than 8 0 percent of them are located in rural areas . . . poverty has worsened. With 60 percent of the total [family] budget devoted to food, one can expect a sub sistence existence for the majority of families (pp. 2; 61).
The Christian Science Monitor (September 19, 1980) added a paradoxical observation:
There is a wide gap between the rich and the poor . . . Mercedes automobiles, walled estates, lavish office structures, and jewelry bedecked women contrast pain fully with enormous slums, serious malnu trition among children, and simple thatched nipa-palm huts with bamboo- stave frames (p. B2).
It is precisely the foregoing conditions that were cited
as one of the legitimate reasons by Ferdinand Marcos, present
Philippine president, to declare martial law on September
21, 1972. He declared that martial law was aimed to "save
the government from the threat of communism, to transform
a sick society of privilege and irresponsibility, and to
eradicate mass poverty" (Marcos, 1976, p. 127). The Presi dent then issued a memorandum declaring that the national
development strategy should "anchor itself on one ruling principle that the interest, objectives, and needs of the
poor working people take precedence over those of the rest"
(Marcos, 1976, p. 70).
One of the avowed objectives, therefore, of the present
Philippine government is the emancipation of the rural poor, particularly the small farmers, from poverty and social in
justice (P.D. 27, 1972). This concern has been translated
into increased allotment of government resources for rural 10 development; thus, the Ten-year Development Plan, 1978-87
(National Economic Development Authority, 1979) listed among other things, the following strategy for development:
To attain self-sufficiency in food and to raise farm incomes, concerted efforts will be direct ed towards the development of rural areas in general. This will require the increased pro vision and improved delivery of essential agri cultural inputs such as, among other things, credit, extension work, marketing facilities, and infrastructure, particularly irrigation • and feeder roads.
The agrarian reform program will be directed towards increasing the productivity and income of agrarian reform beneficiaries. The role of farmer organizations, particularly cooperatives, will be expanded in the areas of marketing and mobilization of rural funds (p. 9).
Statement of the Problem
and Objectives of the Study
In the Philippines, the self-proclaimed emphasis of the government on the small farmers has not been without prob
lems. The differential impact of its programs on this sector of the rural population is well documented (Castillo, 1975 and 1977; International Labor Office, 1974 and 1979). This
finding raises several critical questions. Why do some small
farmers succeed while others do not? Who are the small
farmers who succeed? What are their characteristics? What
factors influence their success? 11
This study was designed to provide answers to the fore going questions. Specifically, the study was designed to—
1. Describe the demographic characteristics of the
small-scale rice farmers in Leyte, Philippines;
2. Identify the political, technical, physical, and
other situational factors that may influence per
formance and perception of small-scale rice farmers
in Leyte, Philippines•
3. Determine the average rice yield, agricultural knowl
edge adoption level, participation level in the acti
vities of the rural development program of Samahang
Nayon, and perception level regarding the future of
the Samahang Nayon organizations of small-scale rice
farmers in Leyte, Philippines; and
4. Investigate the hypotheses that:
4.1 Average rice yield, agricultural knowledge
adoption level, participation level, and per
ception level of small-scale rice farmers in
crease with years in education, years as mem
bers of the Samahang Nayon organization, per
capita monthly income, degree of involvement in
project design and implementation, degree of
technicians' competence, degree of communication
between the technicians and the small-scale
rice farmers, degree of support for the organi
zation by village and town officials, frequency of contact between the technicians and the
farmers, soil fertility, farm to market road
condition, adequacy of flood control, and irri
gation sufficiency.
4.2 Average rice yield, agricultural knowledge adop
tion level, participation level, and perception
level of small-scale rice farmers decrease with
the increase in household size, age, years in
farming, distance to the nearest all weather
road and market, and per capita rice farm area.
4.3 Average rice yield, agricultural knowledge adop
tion level, participation level, and perception
level of small-scale rice farmers who are part-
landowners and landowners, bank-relative cre
ditors, and residing in the western part of
Leyte exceed those who are nonlandowners, usu
rer creditors, noncreditors, and residing in
the eastern part of the province.
Dependent Variables
The four criterion variables of this study represent four outcome measures of the participation of small-scale rice farmers in the Philippine rural development program of Samahang Nayon. These are as follows:
Agricultural knowledge adoption level. This is the
acquisition and use of rice technology by the 13
respondents as indicated by their composite scores
in a checklist of fourteen questions on recommended
rice production practices. Each question assessed
the extent to which a specific technology was
applied by asking the respondents the percentages
of their rice farms in which the technologies were
used; thus, the scores ranged from 1,400 for those
who had applied all the recommended practices in
100 percent of their rice farms to 0 for those who
had not adopted any of the recommended technolo
gies .
Participation level. This represents the respondents'
level of participation in four types of Samahang
Nayon activities during the last twelve months.
The respondents were asked to rate their partici
pation in each of the four types of activities in
a five-point Likert-type scale. A scale of 1 to 5
representing responses from "very poor" to "very
good" was provided for the subjects; thus, possi
ble scores ranged from four to twenty.
Perception level. This indicates the level of the re
spondents' optimism or pessimism regarding the
future of their Samahang Nayon organization as
indicated by their composite scores on a scale of
participants' organizational perception. Using a 14
five-point Likert-type scale, the respondents were
asked to respond to seven statements regarding the
probable events that could happen to their Samahang
Nayon organization within a period of five years.
Composite scores could, therefore, range from
seven, for those who strongly disagreed with all
statements, to thirty-five, for those who strongly
agreed with all statements.
Average rice yield. This indicates the respondents'
gross average rice yield, in cavans per hectare,
during the last two years. Productions affected by
calamities such as typhoon and drought were not
included in the computation.
Independent Variables
There are twenty-five predictor variables in this study classified as follows:
Demographic characteristics. This category includes
the following:
1. Age of respondents in years during last
birthday.
2. Total number of household members
3. Total years in farming
4. Total years as members of the Samahang Nayon
organization 15
5. Per capita monthly income. This was computed by
dividing the total gross monthly income by the
total number of household members.
6. Per capita rice farm area. This represents the
total rice farm area of a respondent divided by
the total number of household members.
7. Status of land ownership. Classification of re
spondents as to ownership of rice farm was as
follows: completely full owner, part owner, or
non-owner. A sub-category of this variable was
the status of those who were operating rice
farms that they did not own. These categories
were tenant, rental, and both tenant and rental.
8. Geographic location of the respondents: eastern
Leyte, for the Waray-speaking respondents resid
ing on the eastern part of the province and
western Leyte, for the Cebuano-speaking respond
ents residing on the western part of the prov
ince .
Technical factors. These factors describe the competence of the two types of government technicians who work with the small-scale farmers. These factors also include the degree of involvement of the respondents on decision 16 making of the Samahang Nayon organization. The follow
ing seven variables were included:
1. Frequency of meeting between the Municipal Devel
opment Officer and the respondents during the
last three months preceding July 1980.
2. Frequency of meeting between the Bureau of Agri
cultural Extension technicians and the respond
ents during the last three months preceding
July 1980.
3. Communication between the Municipal Development
Officer technicians and the members of the or
ganization as rated by the respondents.
4. Respondents' ratings of the technical ability
of the Municipal Development Officers in help
ing the Samahang Nayon solve its organizational
problems.
5. Respondents' ratings of the technical ability
of the Bureau of Agricultural Extension tech
nicians in helping the respondents solve the
problems of their farms.
6. Respondents' ratings of their involvement
during the planning stage of their Samahang
Nayon organization.
7. Respondents' ratings of their involvement in
the implementation state of their Samahang
Nayon organization. Political and other situational factors. Three varia bles were included in this category.
1. Support of local village officials to the
Samahang Nayon organization as rated by the
respondents.
2. Support of town officials to the Samahang Nayon
organization as rated by the respondents.
3. Rice production credit source of respondents.
The categories were none (for those who did not
have any credit), predominantly from rural
banks and relatives, and predominantly from
money lenders or usurers.
Physical factors. This category included four physical factors. Appropriate government technicians were asked to rate each factor in a four-point Likert-type scale.
A scale of 1 to 4 representing responses from "very poor" to "very good" was used.
1. Soil fertility
2. Irrigation sufficiency
3. Farm to market road condition
4. Flood control adequacy
In addition, physical factors include:
5. Distance, in kilometers, of the farm to the
nearest all-weather road.
6. Distance, in kilometers, of the farm to the
nearest town market. 18
Significance of the Study
The Philippines has placed much of its hopes and dreams in the farmer— the "backbone of the nation" and the assumed target of many development efforts of the nation (Castillo,
1977, p. 117). The small farmer is regarded as the crucial and strategic element in the difficult task of rural modern ization. However, while there is general agreement on a vigorous emphasis on small-farmer development, there is lit tle knowledge about the small farmers and the strategy to develop them (Patrick et al., 1975, p. i). Many small-farmer development programs among developing nations have fallen short of expectations (Morss et al., 1975). In the Philip pines, for example, in spite of its large allotment for agri culture which accounts for 25 percent of the total government budget during the last five years, productivity per hectare of its major crops particularly rice, corn, sugarcane, and coconut is still low (Castillo, 1977, p. 924). The Filipino farmer is still poor and is faced with the same age-old prob lems of poverty and underdevelopment; thus, the World Bank, which has loaned the Philippine government billions of dol
lars for agricultural development, raised "serious questions
as to whether its funds are making a dent on poverty in the
Philippines" (Aquino, 1981).
The crucial question,therefore, is understanding who
the small farmers are and what factors influence their 19 performance as agricultural producers. A leading Philippine rural sociologist (Castillo, 1977) echoed this concern when she asked:
But who is the Filipino farmer, the intended beneficiary of policies and programs purport edly designed to influence productivity, employment, and greater equity in his favor (p. 117).
Indeed, a better understanding of the Filipino farmers is of primary significance if effective development strategies are to be devised for them. It was precisely this concern that motivated this researcher to undertake this study.
The choice of rice farmers, who are Samahang Nayon mem bers, as the research subjects and the choice of Leyte as locale of the study offer several logical explanations. The province of Leyte mirrors the problems and characteristics of the rice-growing provinces in the Philippines. It has a very high percentage of low-income families (81 percent); it has low percentage (55 percent) of farm area operated by full owners (Castillo, 1977, pp. 33 and 137). As in other provinces of the country, the rural population is mostly composed of poor small-scale farmers.
On the other hand, the Samahang Nayon is one of the principal government-sponsored programs for increasing the productivity of farmers, particularly small-scale farmers.
In Leyte, as in most other rice-growing provinces of the country, most of the small-scale farmers are members of this association. 20
In sum, the understanding of the small-scale farmers
has implications for bridging current gaps in information
regarding this large segment of the rural population, thus
expanding knowledge and helping establish a basis for more
comprehensive empirical studies on factors affecting the
performance and perception of small-scale farmers. Addition
ally, it should provide timely, helpful information for rural development planners, extension agents, and policy makers.
Delimitation and Scope of the Study
This study is limited to the rice farmers who were mem bers of the Samahang Nayon organization in Leyte, Philip pines. Additionally, it is limited to the investigation of
twenty-five predictor variables and four criterion varia
bles including the relationships between these two sets of variables which are assumed to be linear.
Organization of the Dissertation
This dissertation is divided into five chapters, each with sections and subsections. Briefly, hapter 1 presents
the problem and significance of the study, while hapter 2
discusses the theoretical framework. Chapter 3 explains the methodology from sampling procedure to statistical analysis.
This was followed by hapter 4 which presents the discus
sion of results. The last chapter summarizes the problem,
methodology, and findings. Chapter 5 also includes implica
tions of the research results. CHAPTER II
CONCEPTUAL FRAMEWORK
This chapter summarizes two major competing theories undergirding small-scale farmer programs in developing countries— consensus and conflict (Flinn, Buttell, & Havens,
1975). Major emphasis is placed on the identification of causal variables explaining the behavior and performance of small-scale farmers. A good understanding of these variables is sine qua non to the formulation of development strategies for the rural population.
Small-farmers and peasants are interchangeably used in the discussion. This research takes Firth's (1951, p. 84) definition of peasants: "a system of small producers, with simple technology and equipment, often relying primarily for their subsistence on what they themselves produce."
The Consensus Perspective
The consensus perspective has its roots in the 19th century and, until recently, has been the dominant concep tual perspective in sociology. The writings of Comte, Spen cer, and Durkheim and the work of functional anthropologists like Malinowski and Radcliffe-Brown helped shape the modern consensus perspective (see Turner, 1978; Chambliss, 1973).
21 22
Consensus, social order, integration, social solidarity, equilibrium: these are the key words in the consensus pers pective. Social systems are viewed as being composed of dif ferent interdependent elements which exist in equilibrium.
To maintain this equilibrium, integration of personality systems into the cultural system must occur. Fundamental, therefore, to the consensus perspective is the general notion of social equilibrium and the mechanism that integrates dif ferent levels of social reality to maintain a state of
"homeostasis" (Turner, 1978, p. 37). It emphasizes those aspects of society that are harmonious. Dahrendorf (1959, p. 115) summarized the essential elements of the consensus perspective:
1. Every society is relatively persisting config uration of elements.
2. Every society is well-integrated configura tion of elements.
3. Every element in society contributes to its functioning.
4. Every society rest on the consensus of its members.
Causes of Rural Poverty
In explaining rural poverty, consensus theorists prin cipally look at the life style common to the vast majority of rural poor— their culture. Popularized by Lewis (1969), the life style of the rural poor is described as "composed of a set of behavioral norms deviant from those of the domi nant better off majority which is guided by highly integrated 23
set of attitudes reflective of apathy, defeatism, hopeless-
ness, reliance on chance, and concern with short term grati
fication" (Thomas, 1972, p. 22).
Rogers (1969, p. 115) citing various case studies on peasant agriculture and the arguments of Foster (1962) on
the universality of peasant subculture, listed "ten central
elements" in the subculture of poverty— elements that inhib
it small-scale farmers from adopting new technology:
1. Mutual distrust 2. Low aspiration level 3. Lack of innovativeness 4. Fatalism 5. Lack of deferred gratification 6. Limited time perspective * 7. Familism 8. Dependency upon government authority 9. Localiteness 10. A lack of empathy
Foster (1967, p. 11) summed up the consensus view of peasants. He said, "Time to simmer is an essential part of
this concept of peasant culture, time to integrate diffused
traits and complexes into peasant fabric, to rework them,
to make them harmonious with the functional whole."
Strategy for Rural Development
The basic strategy for rural development under the con
sensus perspective is anchored on the following assumptions
(Thomas, 1972; Boguslaw & Vickers, 1977; Chambliss, 1973;
Rogers & Shoemaker, 1971) :
1. Poverty is a consequence of socialization through
social relationships. 24
2. Motivation is a basic factor inhibiting productivity
of rural peasants.
3. Providing opportunities alone (such as job creation,
profitability and credit) have little effect on
rural productivity since established negative values
and attitudes would prevent the effective utiliza
tion of available opportunities.
4. The rural populations' negative values and attitudes
and their low motivation are the principal determi
nants of their behavior, thus the root cause of
rural poverty. The elimination of rural poverty
depends upon the modification of attitudes and values.
It also depends upon increased motivation through
education.
Illustrative of the consensus view of the small-scale farmers was the dichotomy of early and late adoptors of new agricultural technology offered by Lionberger (1961):
Early Adoptors Late Adoptors
Large farms Small farms High income Low income Take risks Security minded Usually under 5 0 Usually over 60 Actively seeking Complacent or skeptical new ideas Seldom participate in Participate in many formal groups non-local groups
The main thrust, therefore, of the consensus perspec
tive in the development of small-scale farmers is the modi
fication of those characteristics that retard the adoption 25 of new agricultural technology through communications of innovations from sources external to the peasant village
(Rogers, 1969, p. 128). This has been the dominant philoso phy in the establishment of the extension programs in deve loping countries. Thus, over the past decade, farmers' train ing centers, extension programs, and other educational pro grams have been established for the purpose of "modifying" the small-scale farmers1 attitudes, values, and behavior so that they may be innovative and, therefore, productive
(Dalisay, 1974, p. 4; Cuyno et al., 1977, p. 16). Addition ally, there has been increased emphasis in upgrading the quality of extension services offered through (1) use of different channels of communication (television, radio, and newspaper), (2) better training of extension agents, and
(3) use of new technology (movies, video, and computers).
The conceptual framework in the development of small-scale is illustrated in Figure 2.
Criticism
In the late sixties, the Green Revolution program was
launched as a major development strategy under the consensus approach to rural development. The small-scale farmers were using technology considered "dysfunctional and obsolete", thus, low productivity (Cuyno et al., 1977, p. 16). The basic approach was to simplify and packaged new agricultural tech nologies so that these can be easily understood, and, COMMUNICATION
Mass Media
-provides information of widespread interest
Agricultural Agencies
-disseminating information on specific practices -teaching basic principles Small-scale Farmers of fanning -providing special -mutual distrust in technical services personal relations -lack of innovativeness Other Farmers -fatalism Changed attitudes -low aspirational levels & values -social status -lack of deferred -solidarity gratification I -mutual aid -limited time perspective Innovativeness -response -familism -recreation -dependency upon govt. I authority Higher product Commercial Sources -localiteness ivity -lack of empathy -buying S selling materials & equipment -professional services
CHANNELS TARGET CLIENTELE OUTCOMES
Figure 2. Graphic presentation of the consensus approach to small-scale farmer development. 27 therefore, adopted by small-scale farmers. In conjunction with other government services and programs such as easy production credits and cooperative marketing, developing nations embarked on the program with the big hope of not only increasing the production of basic foodstuff but also eliminating rural poverty.
The failure of the Green Revolution as a strategy to effect rural development is well documented (Castillo, 1975
& 1977; Felstehausen, 1973; Thiesenhusen, 1972; Havens &
Flinn, 1973; Diaz & Felstehausen, 1974) . In a number of cases, Green Revolution did not only fail to increase the per capita availability of grains but it also did aggravate the problem of unequal wealth distribution. Studies show that the cost and skill requirements of the new technology are "biased against small producer unless landownership is equally distributed in small parcels and that all peasants have approximately equal access to fertilizers, water, tech nical knowledge, and credit" (Havens & Flinn, 1973, p. 7).
Parelman (1979, p. 249), thus, concluded that the most important constraint to higher yields among small farmers in developing nations is not inadequate technology, but the socio-economic structure which restrict them from using efficiently available resources.
Increasing empirical evidence show that small-scale far mers are, in fact, willing to innovate. These farmers failed to use available technology, not because of their negative 28 values and attitudes, but because they perceived that the profitability that comes with the new technology do not automatically benefit them. The use of the new technology requires greater investment of money and time. Since the new technology is under the "predominant control of those who own most of the land and capital", these people, in fact, become the ultimate beneficiary of Green Revolution, making the small-scale farmers financial situation worst than before (Havens & Flinn, 1973, p. 5).
In view of the inability of the consensus perspective to effect dramatic progress in rural development among developing nations, development planners have looked at other conceptual models for alternative approaches to the solution of rural problems. The succeeding section briefly details the major competing perspective to consensus: conflict.
The Conflict Perspective
The conflict perspective, contrary to some views, is as old as functionalism. It found its inspiration in the works of two German sociologists, Marx and Simmel (Turner,
(1978). Recent conflict theorists like Coser have contributed to the refinement of the theory.
Conflict theorists reject the consensus view of social equilibrium. "They interpret order analysis as a strategy of the ruling group, a reification of their values and motivations, a rationalization for more effective social control" (Horton, undated, p. 704). They view society as an arena for continuing political struggle between social class es or between groups with opposing goals and conception of the world. All social systems, therefore, reveal inequality of resource distribution, the principal cause of conflict, the precursor of reorganization of social system (see Fig ure 3) .
While consensus theorists view tension and strain as dysfunctional, connoting some form of sickness in the system conflict theorists view them as healthy, performing basic preserving functions. They are necessary for peaceable maintenance of relationships; they are a major precursor of social change. Coser summarized the six important functions of conflict (Turner, 1978, p. 161):
1. Conflict permits internal dissention and dissatisfaction to rise to the surface and enables a group to restructure itself or deal with dissatisfaction.
2. Conflicts provides the emergence of new forms of appropriate behavior by surfacing shortcomings.
3. Conflict provides means of ascertaining the strength of current power structures.
4. Conflict works to strengthen the boundaries between groups, bringing out their dis tinctiveness .
5. Conflict creates bond between loosely structured groups, unifying dissent and unrelated elements.
6. Conflict works as a stimulus to reduce stagnation. Conflict may alter society. 30
Deviations from state activities
Specific General Seek or maintain Equilibrium structures ------^ mechanism — ______\ states are types of which of social whole
Consensus Perspective
Inequality All Reveal of resource Causes Conflict Causes Reorganization social ------> distribution ----> which ------* of systems which then Social System
Conflict Perspective
Figure 3. Comparative graphic presentation of consensus and conflict perspective of society (Turner 1978, pp.102 & 185) . 31
Causes of Rural Poverty
Conflict theorists view rural poverty as resulting from
the major problem of social inequality. That is, rural poverty is a manifestation of the greater problem of inequal
ity in wealth, power, prestige, and privilege among different
social classes of people. The focus of analyses is the
social relations that exist between different social classes.
The unequal distribution of economic and political powers among developing nations is well documented (Interna tional Labor Office, 1977). The holders of economic power also control political power; thus, laws and regulations are enacted and enforced to sustain the privileged position of the ruling class.
Gunar Myrdal (1970) who made the inquiry into the pover
ty of nations, observed that while developing countries publicly announce its pursuit for egalitarian ideals, social
and economic inequality seem to be increasing thus pushing
the peasants into a "state of extreme lack of social mobili
ty and severely hampered possibility of competing freely"
(p. 56). He attributed this paradox to the distribution of power and the existing social structure. He described the
situation in Asia;
Fairly independent of the form of govern ment, political power . . . is held nearly everywhere by privileged groups, the first rank including landowners, industrialists, bankers, merchants, and higher military and civilian officials. Under these upper- upper-class groups are . . . the "rural 32
elite" which are composed of peasant land lords, privileged cash tenants, traders, money lenders, officials and so on at the local level (pp. 61-62).
Conflict theorists argue that small-scale farmers, of tentimes, are not free to make decisions for increasing farm productivity and, thus, fail to adopt new technology for in creasing farm efficiency. Two major contributory factors are cited. First, most of them do not own the land they are cul tivating. Second, even if they are landowners, market and credit sources are controlled by the wealthy, making it dif ficult for them to get real benefits from new agricultural technology.
In a situation where small-scale farmers have a high degree of control of the capital, they are most productive and efficient units of production. Mann and Dickinson (1979) called this "petty commodity production" which is character ized by the unity of labor and capital. That is, the small- scale producers own their means of production such as tools, working animals, and land; and together with their families, they provide the labor force for the economic unit.
Marx and Lenin frowned at small-scale peasant agricul ture and claimed that the peasants who produce with their own means of production will either be gradually transformed into small capitalists who also exploit the labor of others or they will suffer the loss of their means of production.
Chayanov (1966), however, believed in the viability of 33 peasant family farms. His hypothesis is based on the concept of the peasant farm as a family labor farm in which the family, as a result of its year's labor, receives a single labor income and weighs its effort against the material re sults obtained. Thus, peasant-owned farms are most efficient units of production since family members are willing to work for long hours and reinvest profit to make their farm more productive.
Strategy for Development
As stated earlier, social and economic inequalities are the main explanatory factors of rural poverty; thus, the principal strategy for effecting development is the removal of the socio-structural barriers that promote these inequal ities (Myrdal, 1970, pp. 56-57). Structural Marxists believe in the radical transformation of existing social system through violent revolution such as in the case of Cuba,
Russia, and the People's Republic of China. Since the ruling class is viewed as without conscience, there can never be a peaceful transfer of power without uprising. A violent over throw of those who are in power by the laboring class is the ultimate prescription for bringing about a genuine socialist change.
Structural non-Marxists, however, believe that meaning ful social reforms for the elimination of social inequities can be achieved within the framework of existing social systems. To achieve their objectives, the use of strikes, 34 boycotts, protest marches, and sit-ins are recommended when regular channels fail.
Since inequalities breed social injustice, the major thrust in the development strategy among conflict theorists is the equitable distribution of wealth through such measures as massive land reform, nationalization of multinational corporations, and progressive taxation. Corollary to econom ic reforms are political reforms that are aimed to diffuse political power through participatory democracy.
The end-result, therefore, of a conflict non-Marxist rural development strategy is production ownership by and active political participation of small-scale farmers. When these conditions are achieved they act as efficient managers and workers of their petty production units (see Figure 4).
Criticism
Outside the People's Republic of China, there has been no widespread demonstrated success on the conflict strategy for rural development, even among communist nations (see
Parelman, 1979). Also, the overthrow of those who are in power does not guarantee the creation of a just and more humane social system. Furthermore, there is the attendant risk of extreme political and social upheaval at the great sacrifice of the very population the revolution is supposed to serve (e.g., Cambodia, Laos, and Iran). HASS ACTION
Social Justice
Protests -equitable distri bution of econo Strikes mic & political power S it-ins Small-scale Farmers Demons trations -Stifled creativity I 6 resourcefulness Higher Creativity Boycotts -/ due to social & Resourcefulness injustice Among Small-scale Other Political Farmers actions I Higher Productivity
STRATEGIES TARGET CLIENTELE OUTCOMES
Figure 4. Graphic presentation of the conflict strategy for small-scale farmer development. 36
Conclusion
Empirical studies on small-scale farmer development projects such as those conducted by Morss (1976), Castillo
(1975), Owen (1974), Watkola (1975), and the United Nations
(1977) and the experiences in Korea (Kim, 1976), India
(Rudramoorthy, 1964), and the People's Republic of China
(Aziz, 1975; Wong, 1973) suggest that the extreme positions
held by two major group of theorists do not hold. The fol
lowing were some of the major agreements on small-farmer
development:
1. The small farmers do quite well to maximize their
economic well-being given the factual and perceived
constraints under which they operate.
2. Their performance is significantly affected by
economic, social, political, cultural, physical,
psychological, and logistical constraints.
3. Generally, the technical production problems are
relatively simple compared to those of socio
economic in nature.
In conclusion, therefore, researches have shown that
the variables explaining small-scale farmers' participation
in rural development are not the monopoly of one theory.
Both personality and structural variables do affect the
performance and outlook of small-scale farmers. This study, therefore, tried to combine the variables that are unique and common to both theories in explaining four outcome measures of small-farmer development program in the Philippines. It was hoped that with an eclectic approach to research, a better empirical understanding of the small-scale farmers will result. CHAPTER III
RESEARCH METHODOLOGY
Population and Sample
The subjects of this study were the rice farmers who were members of the Samahang Nayon organizations in the province of Leyte, Philippines. A random sample of 100 rice farmers was drawn from each of the eastern and western parts of the province. These 200 randomly selected farmers consti tuted the respondents of the study.
Sampling Procedure
The respondents were selected by using a multi-stage sampling procedure that included the following steps:
1. A complete list of all municipalities where rice is
produced in Leyte was made from the latest avail
able data from the provincial offices of the Bureau
of Plant Industry, Bureau of Agricultural Economics,
and Bureau of Agricultural Extension. Thirty-three
rice producing municipalities were identified and
divided into two geographical groups of the prov
ince: eastern (Waray speaking) and western (Cebuano
speaking) municipalities.
38 39
From each geographic location, a random sample of ten municipalities was drawn. These municipalities were as follows:
Eastern Leyte Western Leyte
Abuyog Albuera
Alang-alang Baybay
Burauen Eastern Ormoc
Dagami Hilongos
Javier Hindang
La Paz Inopacan
Mayorga Kananga
Palo Matagob
Pastrana Merida
Tanauan Western Ormoc
In each randomly selected municipality, a complete listing of all villages that met the following criteria was made: one of the major crops grown was rice, and a Samahang Nayon was organizationally func
tioning for the last two years. From each municipal
ity, one rice-producing village with a functional
Samahang Nayon organization was randomly chosen.
From each randomly selected village, a list was made
of all members of the Samahang Nayon organization
that met the following criteria: (1) the member was
engaged in rice productioil and (2) the member had
joined the organization for at least two years. From 40
the list of qualified members, ten respondents were random
ly chosen. A total of 100 farmers were selected from each
geographic location or a total of 200 for the entire prov
ince .
Instrument Development
The major aim of the methodology used was to collect
complete, accurate data from the respondents. To achieve
this objective, an instrument was developed. The subsections
that follow detail the development of the instrument for
this study.
Preliminary Instrument Development
A review of instruments employed in parallel studies
was made. In selecting related instruments, three criteria were observed: the instruments were used to collect data
from farmers in developing countries with low education; the
instruments collected data on variables of interest in this
study; and the instruments had demonstrated validity. From
the review, one important fact emerged. Face-to-face inter
view is the most widely used method of collecting data from
farmers in developing countries. The reasons, of course, are
obvious. The low educational level eliminates the use of
mailed questionnaires; the absence of telephones in rural
areas precludes the use of a telephone interview.
Two instruments in small-scale farmer studies proved
useful in developing the preliminary instrument: the 41
Malaysian study of Zainuddin (1977) and the studies in the
Philippines (Ponce et al., 1975; Ponce et al., 1976). The
interview schedlue used in this research is a modification of the instrument used by the author in his earlier studies
in the Philippines.
To improve the validity of the instrument, it was dis tributed for review, suggestion, and criticism to the writer's dissertation committee and some Filipino social researchers known by the author. The suggestions received were studied and analyzed. Those deemed appropriate were
incorporated. Additionally, the suggestions of the disser tation committee during the proposal approval meeting were also studied and incorporated before the instrument was translated into the local dialects.
The instrument was translated into Cebuano for the re spondents residing in western Leyte and Waray for the east ern part of the province. The initial translations were circulated to selected faculty members of the Visayas State
College of Agriculture (ViSCA) in Leyte, Philippines for
comparability and accuracy in the translation. Suggestions were studied and incorporated before the instrument was pilot tested.
Pilot Testing the Instrument
The Cebuano version was pilot tested in one rice-grow
ing village with a functioning Samahang Nayon organization
in Baybay, Leyte by the Extension Research and Development 42
Division staff of ViSCA. For this purpose, twenty Samahang
Nayon members who were small-scale rice farmers were random
ly chosen and interviewed during the later part of November
to early December of 1979.
Detailed comments and suggestions were solicited from
the two staff members who pilot tested the instrument. These were studied and they formed part of the bases for (1) the
final revision of the instrument in both English and local dialect versions and (2) the final version of the interview er's instruction manual. Additionally, the responses of the pilot respondents were analyzed. Items where a number of answers were either doubtful or incomplete were either revised or eliminated.
To check further the ability of the instrument to ob
tain accurate answers in agricultural knowledge adoption
level and participation level, the Municipal Development
Officer and Bureau of Agricultural Extension technicians assigned to the pilot-tested village, were asked to rate a
few randomly selected respondents on these two dependent variables. Answers of the respondents and those of the raters were compared; these were found to be similar.
Funding constraints prevented the author from pilot
testing the Waray version, i.e., administering the instru ment to pilot respondents to establish a better indicator of
reliability (test-retest method). In fact, the pilot testing of the instrument was done by the ViSCA staff during their 43
off hours for free. However, the Waray dialect is very simi
lar to Cebuano; thus, it was assumed that the problems re vealed during the pretesting of the Cebuano version would
reveal same problems in the Waray version. After all, the
two versions were parallel translations from the same Eng
lish version. Those who helped review the translations
shared this view.
The Final Instrument
The final instrument consisted of a seven-part inter view schedule designed to measure the four dependent and
tWtenty-five predictor variables through face-to-face inter view (see Appendix B). The parts of the interview schedule are as follows:
Part I. Agricultural Knowledge Adoption Level. This
consisted of fourteen questions designed to meas
ure the extent to which the respondents applied
each of the fourteen recommended rice production
practices during the last two cropping seasons
prior to the survey. Extent of adoption was meas
ured by asking the respondents the percentage of
their farm in which a specific technology had been
applied.
Part II. Rice Production. This part of the interview
schedule recorded the respondents' gross rice
production incomes (cavans per hectare), including 44
the areas planted per cropping season during the
last two years prior to the survey. Productions
affected by flood, typhoon, and drought were
excluded.
Part III. Participation Level. This part consisted of
four question designed to measure the extent to
which the respondents participated in the follow
ing activities of their Samahang Nayon organiza
tion during the past twelve months prior to the
survey: attendance in farmer's class, attendance
in meetings, attendance in cooperative activities,
and attendance in social activities. Scoring of
responses were as follows: 1 = very poor (20 per
cent or below attendance); 2 = poor (21 to 40 per
cent attendance); 3 = average (41 to 60 percent
attendance); 4 = good (61 to 80 percent attend
ance) ; and 5 = very good (81 to 100 percent
attendance).
Part IV. Perception Level. This part consisted of seven
statements describing some probable future events
of the Samahang Nayon organizations. Respondents
were asked about their reactions in terms of a
five-point Likert-type scale. Scoring of the re
sponses were as follows: 1 = strongly disagree;
2 = disagree; 3 = neither agree nor disagree;
4 = agree; and 5 = strongly agree. 45
Part V. Gross Income. This portion recorded the gross
incomes of the respondents during the year 1979.
Part IV. Independent Variables. This part consisted of
eighteen questions on demographic, political,
technical, and other situational variables of
interest.
Part VII. Physical Factors. Six questions comprised
this part of the interview schedule. The questions
were designed to assess the following physical
factors: soil fertility, irrigation sufficiency,
flood control, and farm-to-market road condition.
Appropriate government technicians were asked to
rate these items for every respondent using a
Likert-type scale. Additionally, distance of the
respondents' farms to the nearest all-weather
road and distance of the respondents' farms to
the nearest rice markets were also asked from
the respondents.
Data Collection Procedure
The data were obtained through face-to-face interview.
The sections that follow explain the methodological rigor involved to increase the validity and generalizability of the data collected. 46
Selection and Training of Interviewers
The responsibility of hiring and training interviewers rested primarily with the Extension Research and Development
Division staff of ViSCA. Upon the request of the researcher, certain specific procedures were observed to minimize inter viewer's bias.
1. Only interviewers with the following characteristics
were hired: had a college education, had proficiency
in the dialects of the respondents, and had the per
sonality and ability required for conducting effec
tive interviews.
2. With the assistance of the ViSCA staff who field
tested the instrument, the newly-hired interviewers
were given a three-day training program on the
following—
o Objectives of every question in the interview
schedule;
o The process of entering the village and solic
iting the approval of selected respondents;
o The art of asking questions, including probing
through simulations and actual field practice;
and
o The techniques of reviewing completed instru
ments to avoid incomplete returns. 47
Protocol and Data Collection Procedures
The process of actual data collection involved prelimi nary protocol requirements and quality control procedures.
The steps involved were as follows:
1. A commitment to support the study was secured from
the Visayas State College of Agriculture, the prime
beneficiary of the results of this study. This sup
port was in terms of the availability of some col
leagues of the researcher who were staff members of
the Extension Research and Development Division to
(1) supervise and administer the pilot testing and
reproduction of the instrument, (2) select and train
interviewers, (3) randomly select municipalities,
villages, and respondents, (4) get the permission to
conduct the study from relevant offices, and (5) su
pervise and administer the actual data collection.
Once permission was secured, tape-recorded and writ
ten instructions regarding the total mechanics of
the study— from pilot testing the instrument to qual
ity control during actual data collection— were sent
by the researcher to the staff members involved in
the study. Of course, previous to this, some under
standing between the researcher and the concerned
Extension Research and Development Division staff
was already made. 48
2. Permission to conduct the study was secured from the
Provincial Officer of the Ministry of Local Govern
ments and Community Development. Once permission was
granted, site visitations were made to all randomly
selected towns by some members of the Extension
Research and Development Division for the purpose
of—
o Selecting random samples of rice-growing vil
lages and, consequently, random samples of
farmer respondents;
o Informing the Municipal Development Officers
and the Samahang Nayon presidents regarding
the objectives of the study; and
o Asking the Municipal Development Officers and
the Samahang Nayon presidents their coopera
tion in informing the selected respondents
regarding the objectives of the study and the
probable date of the interview.
3. The actual data collection was undertaken in July
and August of 1980 after the following preliminary
steps were completed: drawing the sample of farmer
respondents, revising and reproducing the instrument,
securing the necessary permission, and training the
interviewers.
4. The interviewers were instructed that they should
arrive at the villages at least a day before the 49
start of the interview. This enabled them to gain
familiarity with the villages and to establish
rapport with the barrio folks. Additionally, they
were asked to make courtesy calls to the Barangay
Captains and the Samahang Nayon presidents.
5. During data collection, the interviewers were close
ly supervised by the Extension Research Development
Division staff. Completed instruments were checked
for completeness, accuracy, and missing pages by
the supervising members of the college staff.
6. Completed instruments were edited and sent by air
to the researcher for coding, key punching, and
computer processing.
Data Preparation and Analysis
The completed instruments were examined by the research er for completeness. Data were coded, entered into IBM cod ing forms, and key punched into computer cards. Preliminary runs were made to detect errors in coding, key punching, and entries using the Statistical Analysis System (SAS, 1979) computer program. After errors were corrected, data were analyzed using primarily SAS and, to some extent, BMDP (Bio medical Statistical Package, 1977) computer program.
Descriptive statistics such as frequencies, percentages, means, and standard deviations were computed to summarize the data regarding the small-scale farmers' characteristics. 50
Additionally, simple zero-order correlations were calculated among numeric variables to measure the association between two variables.
Inferential statistics were generated to test the hypotheses of this study, thus making it possible to draw conclusions about the population of the study from the sam ple data (Hays, 1973, p. v). Specifically, the inferential statistics, including their corresponding uses, were as follows:
1. Chi square test of homogeneity. The purpose of the
chi square test was to test whether there was a sig
nificant difference between the eastern and western
respondents in terms of their distribution in the
following nominal variables: status of land owner
ship and credit source of the respondents.
2. Multivariate analysis of variance (MANOVA). This
statistical procedure was used to answer the follow
ing questions:
o Was the population of rice farmers in eastern
Leyte significantly different from the popula
tion of rice farmers from western Leyte in
terms of the twenty-two numeric and four
criterion variables?
o Were the categorical variables— credit source,
land ownership, and province— related to the
four criterion variables? In the first question, one is tempted to use multiple t testing, i.e., series of t tests are per formed to compare the means of one variable at a time by geographic location. Kennedy (1977) warned that this procedure is replete with statistical pit falls. The multiple use of t tests inflates the alpha error. Furthermore, "successive t testing ex ploits the same information several times over, thus violating the independent sample requirement and abusing the machinery of inference" (Kennedy, 1977, p. 82).
MANOVA deals with a vector containing several dependent variables, unlike the analysis of variance
(ANOVA) which deals with a single dependent varia ble. It analyzes vectors of means where each element of the vector is a group's mean for a particular variable; ANOVA, on the other hand, analyzes means of individual variables (Tatsuoka, 1971). Thus,
MANOVA gives an overall group significance, i.e., there is at least two group centroids (group's vec tor of mean scores) that are significantly different from each other.
Canonical variate analysis (CANON). The purpose of using canonical variate analysis was to find some answers to the major question of this study: What are the predictor variables that best explain the 52 average rice yield, agricultural knowledge adoption level, participation level, and perception level of small-scale rice farmers in Leyte, Philippines.
Canonical variate analysis determines the in terrelatedness between two sets of variables— in this study, the relationship between the set of twenty-two numeric predictor variables and the set of four continuous criterion variables.
The basic strategy of canonical variate analy sis is to "devise a linear combination from each of the sets of variables in such a way that the corre lation between the two linear combinations is max imized" (Warwick, 1975, p. 517). First, the maximum correlation between the two linear combinations is sought, then additional pairs of independent linear combinations are calculated.
For each pair of linear combinations, four im portant types of information are produced by canon ical variate analysis: the canonical variates, the canonical correlations between them, the structure coefficients, and the redundancies. Since the canon ical variate for each set of variables displays the coefficients that reflect the importance of the original variables in the subset in forming the variate, it makes it possible for the researcher to understand and interpret the structure of the 53 relationship between the two sets of variables (War wick, 1975, p. 517; Levine, 1977). Thus, by examin ing the structure of relationships between the two sets of variables, the important variables in both sets that best explains a particular significant canonical correlation are, therefore, identified.
Additionally, eigenvalues were also generated for better understanding of the amount of variance in one canonical variate that is accounted for by the other canonical variate.
Ten separate canonical variate analysis proce dures were performed— five for each population of farmers according to geographic location. For each geographic location, therefore, a canonical variate analysis between the set of twenty-two numeric and the set of four criterion variables was first per formed. As a method of simplifying the complex mul tivariate relationships identified by the first canonical analysis, four reduced canonical analyses were also made. This approach was logical because the MANOVA procedure showed that the eastern and western groups of farmers differed significantly in a number of variables; thus, they can be considered as two distinct populations. This findings confirmed the observation of the researcher who had been a resident of Leyte for six years. With this procedure, 54 there is a better understanding of the relationship between the two sets of variables under slightly different cultural settings.
In summary, the canonical variate analysis was utilized in this study to provide information con cerning (Levine, 1977, p. 12):
1. The nature of the links or patterns of interdependency that join the two variable sets;
2. The number of (statistically signifi cant) links between the sets; and
3. The extent to which the variance in one set is conditional upon or re dundant given the other set.
The BMD computer program was used for canonical variate analysis since it generated the kind of in formation needed by the researcher for meaningful interpretation. CHAPTER IV
FINDINGS AND DISCUSSION
This chapter contains the findings of the study. Each geographic location is separately presented and discussed for better understanding of the variables being investigat ed. This chapter is divided into two parts. Part 1 presents the profile of the two groups of respondents. Part 2 dis cusses the relationships between the set of predictor variables and the set of criterion variables.
The Respondents: Two Profiles
One of the major objectives of this study was to des cribe the characteristics of the small-scale rice farmers in
Leyte, Philippines who were also members of the major gov ernment program for small-farmer development— the Samahang
Nayon. The profiles include the demographic characteristics and the following three categories of situational variables that are likely to influence the performance and perception of the respondents: technical, political, and physical factors. In addition, the profiles also include the distri bution of the respondents by agricultural knowledge adoption level, average rice yield, participation in the activities of the Samahang Nayon organization, and perception regarding the future of the Samahang Nayon organizations— the four
55 56 criterion variables of this study. The two subsections that follow present the profiles of the two groups of small-scale rice farmers in Leyte.
The Small-scale Rice Farmers of Eastern Leyte
The respondents from eastern Leyte were randomly drawn from the Waray-speaking, rice-producing municipalities of the province.
Demographic characteristics. As shown in Table 1, the majority (76 percent) of the eastern Leyte respondents were over forty years old, with forty-nine as the mean age for the group. They belong to households with an average of seven members. On the average, they had five years of educa tion, twenty-four years of farming experience, six years as members of the Samahang Nayon organization, and with a per capita monthly income of ^101 (US $14). Additionally, they were producing rice on an average area of less than one- fifth hectare (.4 acre) per family member on rice lands that almost half (49 percent) of them did not completely own.
Only a little over one-third completely owned their rice- fields. Seventy-three percent of those who were completely or partially non-owners were tenants, an arrangement where a landowner receives a fixed percentage of the gross produce.
Technical factors. This category of independent varia bles included seven items ranging from the technical expert ise of government technicians to frequency of meetings 57
TABLE 1
DISTRIBUTION OF RESPONDENTS
BY DEMOGRAPHIC VARIABLES AND GEOGRAPHIC LOCATION
Items East West Total N N N %
30 below 3 4 7 3.5 31 40 21 25 46 23.0 41 50 37 29 66 33.0 51 60 26 25 51 25.5 61 above 13 17 30 15.0 Total 100 100 200 100.0
Mean 49.12 48.20 48.66 S.D. 11.06 11.82 11.43
Education in Years
0 8 15 23 11.5 1-2 8 19 27 13.5 3 - 6 65 54 119 59.5 7-10 16 8 24 12.0 11 & above 3 4 7 3.5 Total 100 100 200 100.0
Mean 4.98 3.82 4.40 S.D. 2.73 2.88 2.86
Total Household Members
3 & below 8 17 12.5 4 - 6 45 45 45.0 7 - 9 33 31 32.0 10 - 12 10 7 8.5 13 & above 4 0 2.0 Total 100 100 100.0
Mean 6.65 5.89 6.27 S.D. 2.63 2.25 2.47 58
TABLE 1 (continued)
Items East West Total N N N %
Years in Farming
5 & below 7 3 10 5.0 6-15 23 16 39 19.5 16 - 25 25 28 53 26.5 26 - 35 27 27 54 27.0 36 & above 18 26 44 22 .0 Total 100 100 200 100.0
Mean 24.37 28.00 26.19 S.D. 13.44 13.06 13.34
Years as Members of SN
2 - 3 7 7 14 7.0 4 - 5 14 6 20 10.0 6 - 7 77 87 164 82.0 8 & above 2 0 2 1.0 Total 100 loo 200 100.0
Mean 6.24 6.23 6.24 S.D. 1.35 1.13 1.24
Per Capita Monthly Income (peso)
150 & below 80 71 151 75.5 151 - 200 9 11 20 10.0 201 - 250 3 6 9 4.5 251 - 300 3 4 7 3.5 301 - 350 0 2 2 1.0 351 - 400 3 1 4 2.0 401 & above 2 5 7 3.5 Total 100 100 200 100.00
Mean 100.98 136.13 118.56 S.D. 103.98 166.25 139.43 59
TABLE 1 (continued)
Items East West Total N N N %
Per Capita Rice Farm Area (hectare)
.100 & below 35 30 65 32.5 .101 - .200 32 29 61 30.5 .201 - .300 18 14 32 16.0 .301 - .400 8 13 21 10.5 .401 - .500 4 8 12 6.0 .501 & above 3 6 9 4.5 Total 100 100 200 100.0
Mean .18 .24 .21 S.D. .14 .26 .21
Status of Rice Land Ownership
Completely non-owner 49 72 121 60.5 Partly owner 14 15 29 14.5 Full owner 37 13 50 25.0 Total 100 100 200 100.0
Chi square = 15.93; p = .0003
Land Arrangement of Non-owners
Leasehold 17 52 69 46.6 (rental system) (27. 4%) (60 .5%) Tenant 45 34 79 53.4 (share system) (72. 6%) (39 .5%) Total 62 86 148 100.00
Chi square = 15.81; p = .0001 60
(Table 2). The Municipal Development Officers, the techni cians assigned by the government to provide technical as sistance to Samahang Nayon organizations relative to organi zational problems, were rated by 53 percent of the respond ents as average in technical ability (neither good nor poor). Forty-three percent rated them as good. However, when the respondents were asked on the frequency they met with these technicians during the last three months prior to the survey, 85 percent answered none. Although there was virtu ally no contact between the Municipal Development Officers and the clientele in the few months prior to the survey, al most one-half (47 percent) of the respondents felt that com munication between the technicians and the clientele was good. When respondents were asked further regarding their involvement in the implementation of the programs of their organizations, 58 percent claimed that they were either rarely or never involved; although only 35 percent gave the same rating when asked about their involvement during the organizations' initiation stage. The mean for participants involvement was 1.83 (rarely involved), while the mean for participants' involvement during the planning stage was 2.83
(sometimes involved).
The Bureau of Agricultural Extension technicians, popu larly called by the farmers as BAEX technicians, work indi vidually with small-scale rice farmers regarding farm prob lems. In comparison with the Municipal Development Officers, 61
TABLE 2
DISTRIBUTION OF RESPONDENTS
BY TECHNICAL FACTORS AND GEOGRAPHIC LOCATION
Items East West Total N N N %
Frequency of Meetings Between MDOs & Respodents (April-June 1980)
0 85 95 180 90.0 1 - 2 15 5 20 10.0 Total 100 100 200 100.0
• Mean .13 .13 .13 S.D. .36 1.20 .88
Technical Ability of MDOs as Rated by Respodents
1 - Very poor 3 2 5 2.5 2 - Poor 1 25 26 13.0 3 - Neither good nor poor 53 52 105 52.5 4 - Good 43 18 61 30.5 5 - Very good 0 3 3 1.5 Total 100 100 200 100.0
Mean 3.25 2.88 3.07 S.D. .76 .82 .82
Communication Between MDOs & Respondents as Rated by Respodents
1 - Very poor 2 2 4 2.0 2 - Poor 4 28 32 16.0 3 - Neither good nor poor 45 50 95 47.5 4 - Good 47 15 62 31.0 5 - Very good 2 5 7 3.5 Total 100 100 200 100.0
Mean 3.25 2.88 3.12 S.D. .79 .86 .86 62
TABLE 2 (continued)
Items East West Total N NN %
Respondents' Planning Involvement
0 - Never 3 3 6 3.0 1 - Very rarely 7 16 23 11.5 2 - Rarely 25 27 52 26.0 3 - Sometimes 32 37 69 34.5 4 - All the time 33 17 50 25.0 Total 100 100 200 100.0
Mean 2.85 2.49 2. 67 S.D. 1.06 1.05 1 .07
Respondents 1 Participation Involvement
0 - Never 29 0 29 14.5 1 - Very rarely 18 19 37 18.5 2 - Rarely 11 36 47 23.5 3 - Sometimes 25 28 53 26.5 4 - All the time 17 17 34 17.0 Total 100 100 200 100.0
Mean 1.83 2.43 2. 13 S.D. 1.50 .99 1 .30 Frequency of Meetings Between BAEXs & Respondents (April-June 1980)
0 51 42 93 46.5 1 - 2 21 4 25 12.5 3 - 4 21 36 57 28.5 5 & above 7 18 25 12.5 Total loo 100 200 100.0
. Mean 1.67 2.65 2.16 S.D. 3.55 3.52 3.57 63
TABLE 2 (continued)
Items East West Total N N N
Technical Ability of BAEXs as Rated by Respondents
1 - Very poor 1 5 6 3.0 2 - Poor 3 23 26 13.0 3 - Neither goc nor poor 26 33 59 29.5 4 - Good 61 27 88 44.0 5 - Very good 9 12 21 10.5 Total 100 100 200 100.0
Mean 3.72 3.17 3.45 S .D. .73 1.08 .96 64 the BAEX technicians were rated more favorably in technical ability, with 61 percent of the respondents rating them as good and 9 percent, very good. Additionally, they met more often with their clientele, with 49 percent of the respond ents claiming a meeting of at least once during the last three months prior to the survey date.
Political factors. Table 3 shows the distribution of respondents by political factors and geographic location.
Two-fifths of the respondents rated the support of the vil lage officials for the Samahang Nayon organization as aver age (neither good nor poor), and almost one-fourth rated the village officials as good. The respondents' rating of the support given by town officials parallels their rating for village officials, although, the former were rated a little more favorably. Twenty-seven percent rated the support of town officials as good. The mean rating for town officials was 3.04; the mean rating for local officials, 2.86.
In terms of credit source, almost one-half (47 percent) claimed that major source of credit for production inputs like fertilizers and chemicals were usurers, while one- fourth indicated rural banks and relatives.
Physical factors. This category of predictor variables includes six factors that directly or indirectly affect the criterion variables (Table 4). The BAEX technicians indicat ed that while the majority (98 percent) of the rice fields had either good or very good soil fertility, 77 percent of 65
TABLE 3
DISTRIBUTION OF RESPONDENTS
BY POLITICAL FACTORS AND GEOGRAPHIC LOCATION
Items East West Total N N N %
Support of Village Officials as Rated by Respondents
1 - Very poor 3 0 3 1.5 2 - Poor 33 7 40 20.0 3 - Neither good nor poor 41 30 71 35.5 4 - Good 22 42 64 32.0 5 - Very good 1 21 22 11.0 Total 100 100 200 100.0
Mean 2.86 3.78 3. 32 S.D. .84 .86 • 97
Support of Town Officials as Rated by Respondents
1 - Very poor 4 3 7 3.5 2 - Poor 23 18 41 20.5 3 - Neither good nor poor 42 51 93 46.5 4 - Good 27 26 53 26.5 5 - Very good 4 2 6 3.0 Total 100 100 200 100.0
Mean 3.04 3.06 3. 05 S.D. .91 .80 86
Credit Source
None 28 43 71 35.5 Predominantly from rural banks & relatives 25 32 57 28.5 Predominantly from money lenders (loan sharks) 47 25 72 36.0 Total 100 100 200 100.0
Chi square = 10.751; p = .005 66
TABLE 4
DISTRIBUTION OF RESPONDENTS
BY PHYSICAL FACTORS AND GEOGRAPHIC LOCATION
Items East West Total N N N %
Soil Fertility
1 - Very poor 0 0 0 0.0 2 - Poor 2 17 19 9.5 3 - Good 78 83 161 80.5 4 - Very good 20 0 20 10.0 Total 100 100 200 100.0
Mean 3.18 2.83 3. 01 S.D. .44 .38 • 44
Irrigation Sufficiency
1 - Very poor 43 5 48 24.0 2 - Poor 34 62 96 48 .0 3 - Good 23 33 56 28.0 4 - Very good 0 0 0 0.0 Total 100 100 200 100.0
Mean 1.80 2.28 2. 04 S.D. .79 .55 • 72
Farm to Market Road Condition
1 - Very poor 20 0 20 10.0 2 - Poor 39 58 97 48.5 3 - Good 21 42 63 31.5 4 - Very good 20 0 20 10.0 Total 100 100 100 100.0
Mean 2.41 2.42 2. 42 1.03 .49 .80 67
TABLE 4 (continued)
Items East West Total N N N %
Flood Control Adequacy
1 - Very poor 0 2 2 1.0 2 - Poor 38 47 85 42.5 3 - Good 0 0 0 0.0 4 - Very good 62 51 113 56.5 Total 100 100 200 100.0
Mean 2.62 2.44 2. 53 S.D. .49 .51 50
Di-stance of Farm to Nearest Town Market (kilometers)
1.00 or less 24 0 24 12.0 1.01 - 2.50 30 2 32 16.0 2.51 - 4.00 8 37 45 22.5 4.01 - 5.50 6 10 16 8.0 5.51 & above 32 51 83 41.5 Total 100 100 200 100.0
Mean 3.62 8.32 5.97 S.D. 2.72 6.02 5. 22
Distance of Farm to Nearest All-Weather Road (kilometers)
.50 or less 72 28 100 50.0 .51 - 1.00 9 10 19 9.5 1.01 - 1.50 5 5 10 5.0 1.51 - 2.00 6 40 46 23.0 2.01 & above 8 17 25 12.5 Total 100 100 200 100.0
Mean .63 1.47 1 .05 S.D. .95 1.01 1.07 68 them had poor and very poor irrigation. This means that for over three-fourths of the farms in the east, there was in adequate available irrigation water— a very important factor in the cultivation of lowland rice. However, in terms of flood control adequacy, 62 percent of the farms were rated by the technicians as very good.
Farm-to-market road condition for 69 percent of the re spondents were either poor or very poor, with 72 percent of them claiming to be within one-half kilometer (.3 mile) to the nearest all-weather road and 2.5 kilometers (1.6 miles) to the nearest town market, the commercial center where far mers usually sell their products and buy production inputs.
Criterion variables. The four criterion variables are outcome measures of participating in the rural development program of the Samahang Nayon. Three of these deal with the performance of the small-scale rice farmers as members of the Samahang Nayon organization: agricultural knowledge adoption level, average rice yield, and participation level in the activities of the organization. The fourth dependent variable, perception level, indicates the respondents' degree of optimism or pessimism regarding their organization
(Table 5). On the average, 35 percent of the fourteen recom mended practices were adopted by the respondents; however, one-fourth of the respondents adopted only 20 percent and below. The average rice yield was 48 cavans per hectare, 69
TABLE 5
DISTRIBUTION OF RESPONDENTS
BY CRITERION VARIABLES AND GEOGRAPHIC LOCATION
Items East West Total NN N %
Agricultural Knowledge Adoption Level (average % of 14 items)
20 & below 24 0 24 12.0 21 - 40 33 19 52 26.0 41 - 60 33 53 86 43.0 61 - 80 10 22 32 16.0 81 - 100 0 6 6 3.0 Total 100 100 200 100.0
Mean 35.36 54.00 44.68 S.D. 18.52 15.61 19.47
Average Rice Yield (Cavans per hectare)
35 &below 34 16 50 25.0 36 - 50 28 17 45 22.5 51 - 65 20 24 44 22.0 66 - 80 9 21 30 15.0 81 - 95 3 11 14 7.0 96 & above 6 11 17 8.5 Total 100 100 200 100.0
Mean 47.85 62.86 55.37 S.D. 27.05 26.75 27.87
Participation Level (average of 4 items)
Very poor (1.4 & below) 0 0 0 0.0 Poor (1.45 to 2.44) 2 1 3 1.5 Neither good nor poor (2.45 to 3.44) 27 24 51 25.5 Good (3.45 to 4.44) 60 52 112 56.0 Very Good (4.45 & above) 11 23 34 17.0 Total 100 100 200 100.0
Mean 3.64 3.84 3.75 S.D. .56 .70 .64 70
TABLE 5 (continued)
Items East West Total N N N %
Perception Level- (average of 7 items)
Very poor (1.4 & below) 0 0 0 0.0 Poor (1.45 to 2.44) 12 3 15 7.5 Neither good nor poor (2.45 to 3.44) 62 75 137 68.5 Good (3.45 to 4.44) 26 22 48 24.0 Very good (4.45 & above) 0 0 0 0.0 Total 100 100 200 100.0
Mean 3.12 3.15 3.13 S.D. .53 .41 .47 71 although 34 percent of the respondents were producing only
35 cavans and below.
In terms of the respondents' participation in the acti vities of the Samahang Nayon organization, 71 percent consi dered themselves as either good or very good. Nobody rated themselves as very poor. For the respondents' perception level regarding the future of their Samahang Nayon organiza tion, the majority (62 percent) were uncertain (neither good nor poor).
The Small-scale Rice Farmers of Western Leyte
The farmer-respondents who were members of the Samahang
Nayon organizations in western Leyte were randomly drawn from the Cebuano-speaking municipalities of the province facing the Camotes Sea.
Demographic characteristics. Like the respondents from eastern Leyte, almost three-fourth (71 percent) of the west ern respondents were over forty years old, with forty-eight years as the mean for the group (Table 1). On the average, they had households with six members, four years of formal schooling, twenty-eight years of farming experience, six years as members of the Samahang Nayon organization, and a per capita monthly income of ?136 (US $19) . As shown in
Table 1, they were cultivating rice on an average area of one-fourth hectare (.5 acre) per family member on fields where 72 percent of them did not completely own. Only 13 72 percent owned their ricefields. For those who were operating lands they did not own, 61 percent were leaseholders or on rent system while 40 percent were tenants— an arrangement where the owner receives a fixed percentage of the gross rice produce regardless of the production level.
Technical factors. The distribution of respondents by technical factors and geographic location is presented in
Table 2. The Municipal Development Officers assigned to the west were rated by 79 percent of the respondents as average
(neither good nor poor) or below average (poor to very poor). A little less than one-fifth (18 percent) rated them as good. As to frequency of meeting, however, 95 percent claimed that they never met with their Municipal Development
Officers during the last three months prior to the survey.
Thus, it was not surprising that one-half of the respondents claimed that communication between the technicians and the clientele was average (neither good nor poor) while 28 per cent claimed it to be poor. In terms of the respondents in volvement during the project initiation, 43 percent claimed they were either rarely or very rarely involved. Only 17 percent claimed that they were involved at all times. When asked of their involvement in the implementation of the pro grams for their organization, the rating was slightly lower with 55 percent claiming that they were either rarely or very rarely involved. 73
As stated earlier, the Bureau of Agricultural Extension technicians are charged with helping the small-scale rice farmers solve their farm problems. Like the east, these technicians were rated more favorably than the Municipal
Development Officers. One-third rated them average; 39 per cent, either good or very good. In addition, these techni cians met more often with their clientele during the last three months prior to the survey. Forty-eight percent claim ed that they met at least once during the period, while 42 percent claimed they did not have any meeting with the technicians at all.
Political factors. The support of village officials for the Samahang Nayon organization was seen as generally better than the support of town officials (Table 3). Sixty-three percent rated the support of village officials as either good or very good, while only percent gave the same rating for town officials. The mean rating for support of village officials was 3.78 (good); town officials, 3.04 (neither good nor poor) .
As for credit source, 43 percent did not have any pro duction credit while one-fourth borrowed from usurers. The remaining respondents obtained their production loans from banks and relatives.
Physical factors. The distribution of the respondents by physical factors and geographic location is shown in
Table 4. Eighty-three percent of the respondents' farms 74 were rated by the technicians as good; however, 67 percent had poor irrigation. In terms of flood control, almost one- half (49 percent) was rated either as poor or very poor.
Fifty-one percent was considered as having very good flood control.
The farm-to-market road condition was considered poor in a little over one-half (58 percent) of the villages. Dis tance of farm to the nearest town market was over 2.5 kilo meters (1.6 miles) for 98 percent of the respondents, al though 43 percent claimed that they are within 1.5 kilome ters (1 mile) to the nearest all-weather road.
Criterion variables. The distribution of respondents by criterion variables and geographic location is shown in
Table 5. On the average, 54 percent of the fourteen recom mended rice production practices were adopted by the re spondents. A little less than one-third (28 percent) were applying more than 80 percent. Average rice yield was fifty- four cavans per hectare with 46 percent of the respondents producing more than sixty-five cavans per hectare.
In terms of the respondents' participation in the acti vities of the Samahang Nayon organization, 75 percent rated themselves as either good or very good. In terms, however, of their perception of the future of their organization, the same percentage felt uncertain (neither good nor poor). 75
Summary: A Comparison of the Two Populations
To determine whether the eastern and western respond ents differ from each other in terms of the predictor and criterion variables, two statistical procedures were per formed: chi square for determining whether the distribution of the categorical variables in the two geographic locations were significantly different and multivariate analysis of variance (MANOVA) for testing whether the two groups of far mers differ from each in terms of the twenty-two numeric predictor and four criterion variables.
Results of the chi square tests. The distribution of
farmers according to land ownership and the distribution of non-land owners according to land arrangement were signifi cantly different (£ = .0003) in the two geographic locations
(Table 1). In the west, almost three-fourth (72 percent) of the farmers were completely non-owners. Only 13 percent were
full land owners. In the east, the percentage of non-owners was lower (49 percent) while the percentage of full land owners was higher (37 percent). In terms of land arrangement
of those who were cultivating on land they did not own,
those in the west were mostly leaseholders (60.5 percent) while those in the east were mostly tenants (72.6 percent).
The distribution of farmers according to credit source was also significantly different (£ = .005) in the two geog
raphic locations (Table 3). In the west, almost one-half of
the farmers (43 percent) borrowed from the rural banks. In 76 the east, however, almost the same percentage (47 percent) borrowed from private money lenders.
Results of the MANOVA. The results of the MANOVA are summarized in Table 6 indicating a significant difference on the centroids of the two groups of respondents (p = .0001).
TABLE 6
MULTIVARIATE ANALYSIS OF VARIANCE OF
ALL NUMERIC VARIABLES BY GEOGRAPHIC LOCATION
Source S M N Hotelling- Signifi- Lawley Trace cance Level
Geographic Location 1 12.0 85.5 2.888 .0001
Where: P = No. of dependent S = MIN (P,Q) variables M = .5 (ABS(P-Q)-l) Q = Hypothesis DF N = .5 (NE-P-1) NE = DF of E E = Error SS & CP matrix
As a follow-up procedure for determining the specific variables that contribute to the difference between the two centroids, multiple comparisons were performed with the use of Bonferoni Inequality (Timn, 1975) . The numeric variables whose means were significantly different between the two groups of respondents are summarized in Table 7. It is shown that the two groups differed in nine predictor variables.
Compared to western respondents, the eastern respondents had: (1) rated higher their Municipal Development Officers and Bureau of Agricultural Extension technicians in technical 77
TABLE 7
SUMMARY OF SIGNIFICANT MULTIPLE COMPARISONS (p < .02)
Variable Conclusion
Technical
1. Technical Ability, Eastern MDOs were rated higher MDO in technical ability than western MDOs.
2. Communication Flow, Eastern MDOs had more frequent MDO and Farmers communication with their clientele than western MDOs.
3. Participation Western farmers had greater Involvement degree of participation in the operation of their Samahang Nayon organizations than east ern farmers.
4. Technical Ability, Eastern BAEX technicians were BAEX Technician rated higher in technical abil ity than western BAEX techni cians .
Political
1. Support of Village Western village officials were Officials rated higher in terms of their support for the SN organiza tions than eastern village officials.
Physical
1. Soil Fertility Eastern rice farms were rated higher in soil fertility than western rice farms.
2. Irrigation Western rice farms were rated higher in irrigation sufficien cy than eastern rice farms. 78
TABLE 7 (continued)
Variable Conclusion
Physical
3. Distance of Farm to Western rice farmers were Nearest Town Market located farther from the near est town market than eastern rice farmers.
4. Distance of Farm to Western rice farmers were Nearest All-Weather located farther from the near Road est all-weather road than eastern rice farmers.
Criterion
1. Agricultural Knowledge Western rice farmers had high Adoption Level er level of agricultural knowledge adoption in rice technology than eastern rice farmers.
2. Average Rice Yield Western farmers had higher rice yields than eastern farmers. 79 ability, (2) more frequent communication with their Munici pal Development Officers, (3) more fertile rice farms, and
(4) closer access to all-weather roads and town markets, being nearer in distance. In contrast, the western respond ents had: (1) higher participation in the activities of their Samahang Nayon organizations, (2) greater degree of support in their organization from their village officials, and (3) better irrigation for their ricefields.
Comparing the two groups of respondents on the four criterion variables, the two groups differed significantly in two variables. Western rice farmers had a higher agricul tural knowledge adoption level and a higher average rice yield than eastern rice farmers.
Relationships Between the Sets of
Predictor Variables and the Set of Criterion Variables
This section deals with two multivariate statistical procedures for testing multivariate relationships. Specifi cally, multivariate analysis of variance (MANOVA) was per formed to test research hypothesis 4.3. That is, it was em ployed to test the relationships between the four categori cal predictor variables and the four dependent variables.
Canonical variate analysis was used to determine the rela tionship between the set of twenty-two numeric predictor variables and the set of four criterion variables and as a 80 test, therefore, of research hypotheses 4.1 and 4.2. The subsections that follow explain the procedures and results.
Results of the MANOVA
A three-factor multivariate analysis of variance was used to examine the relationship between the following two sets of variables:
Predictor Variable Set Criterion Variable Set
1. Geographic Location 1. Agricultural Knowledge (2 levels) Adoption Level
2. Land Ownership (2 levels) 2. Average Rice Yield
3. Credit Source (3 levels) 3. Participation Level
4. Perception Level
The results of the MANOVA are summarized in Table 8. It is shown that significant effects were found for geographic location, geographic location by land ownership, and land ownership by credit source. To investigate more closely these significant effects on each of the four criterion variables, a three-way univariate analysis of variance
(ANOVA) was performed for each criterion variable. This pro cedure sometimes, termed as step-down ANOVA, is recommended as a follow-up procedure when the multivariate alpha is sig nificant (Fink & Kosecoff, 1978). Results of the ANOVAs are
summarized in Table 9.
It is seen that geographic location had significant main effects on all criterion variables except perception
level. At the same time, it interacted with land ownership 81
TABLE 8 MULTIVARIATE ANALYSIS OF VARIANCE OF THE CRITERION VARIABLES BY GEOGRAPHIC LOCATION, LAND OWNERSHIP, AND CREDIT SOURCE
Source S M N Hotelling- Significant Lawley Trace Level
Geographic 1 1.0 91.5 .250 .0001 Location Land Ownership 1 1.0 91.5 .023 .3811 Credit Source 2 .5 91.5 .074 .0939 Geographic 1 1.0 91.5 .052 .0504 Location X Land Ownership
Geographic 2 . .5 91.5 .066 .1520 Location X Credit Source Land Ownership 2 .5 91.5 .113 .0091 X Credit Source Geographic 2 .5 91.5 .0304 .6904 Location X Land Ownership X Credit Source
where: P = No. of dependent S = MIN(P,Q) variables M = .5(ABS(P-Q)-1 Q = Hypothesis DF N = ,5(NE-P-1) NE = DF of E E = Error SS & CP matrix TABLE 9 SUMMARY OF ONE-WAY ANALYSES OF VARIANCE OF EACH CRITERION VARIABLE BY GEOGRAPHIC LOCATION, LAND OWNERSHIP, AND CREDIT SOURCE
Source df Significance Level
Agricultural Knowledge Adoption Level
Geographic Location 1 60.05 .0001 Average Rice Yield
Geographic Location 1 15,66 .0005 Land Ownership X Credit Source 2 3.01 .0443 Participation Level Geographic Location 1 5.49 .0370 Land Ownership X Credit Source 2 4.52 .0433 Perception Level Geographic Location X Land Ownership 5.41 .0022 Land Ownership X Credit Source 5.48 .0054 83
producing an effect on perception levels. Additionally, land
ownership and credit source had interaction effects on all
criterion variables except agricultural knowledge adoption
level. These significant interaction effects on the dependent
variables were further investigated by plotting the cell means. Since the cells had unequal sizes, least square means
including their probabilities and test for multiple compari
sons were generated with the use of the SAS computer program
(see Appendix A).
The disordinal interaction effects of land ownership
and credit source are in Figures 5 and 6. Examination of the
graphs and the multiple comparisons for least square means
showed that:
1. Credit source did not make a difference on the aver
age rice yield of small-scale farmers who were part-
owners and owners. Among the non-land owners, how
ever, the average rice yield of the non-creditors was
significantly higher than the usurer creditors. In
the same group, there was no significant difference
in the average rice yield between usurer borrowers
and bank-relative borrowers or between non-creditors
and bank-relative creditors.
Comparing land ownership by credit source on
average rice yield, it is shown that significant
difference existed only between non-land owners who 84
65 A where: V E A “ Non-owners R 60 B - Part-owners & A Owners G E 0 - No credit 1 - Banks & R 55 Relatives I 2 - Usurers C E
Y 50 I E L D AB LAND OWNERSHIP
LAND OWNERSHIP LAND OWNERSHIP
Figure 5. The interaction effects of land ownership and credit source on average rice yield, participation level, and perception level. 85
CREDIT SOURCE CREDIT 2 ■ • • ■ ■ ■ 23 22 21 20 1 R P E I L L C E P 0 E E T N V CREDIT SOURCE CREDIT B 60 55 50 65 45
R I E A G E R E I A V Y E L D C
CREDIT SOURCE CREDIT
and perceptionand level. source source on average rice yield, participation level,
Owners t Part-owners 6 Banks Relatives Usurers Non-owners No credit - - 14 16 0 - 1 2 13 15 B - A- Figure 6. Figure 6. The interaction effects of land ownershipby credit where 0. < (6 ft H U H a. < ft w O Z JU>W^ borrowed from relatives and friends and part-owners
or owners who borrowed from usurers.
The disordinal interaction effects of credit source
by land ownership on participation level show that
among non-land owners, the participation scores of
those without credit were significantly higher than
those who borrowed from loan sharks. Among part-
owners and owners, however, those who borrowed from
banks and relatives had significantly higher parti
cipation than those who borrowed from usurers, and
those without credit had also higher participation
score than those who borrowed from usurers.
Comparing the participation level of different creditors by land ownership, it was found that the
small-scale rice farmers who were either part-owners
or owners and those whose main source of credit were
banks and relatives had significantly higher degree
of participation than the small-scale farmers who
did not own their land regardless of their credit
source.
The disordinal interaction effects of credit source
by land ownership on perception level showed that
among non-land owners, those without credit had sig
nificantly higher perception scores than either
bank-relative and usurer creditors. Within the same
group, bank-relative creditors had higher perception 87
scores than usurer creditors. However, among part-
owners and owners, significant difference existed
only between bank-relative creditors and those with
out credit.
Comparing the perception level of different
creditors by land ownership, significant differences
in scores were found between (1) part-owners and
owners without credit and non-land owners who either
did not have any credit or who borrowed from banks
and relatives and (2) part-owners and owners who had
credits from banks and relatives and from non-land
owners whose credit sources were the usurers.
The disordinal interaction effect of geographic loca tion and land ownership on perception level is shown in
Figure 7. In the east, non-land owners had significantly higher scores than part-owners or owners. In the west, how ever, the reverse was true.
Summary of the MANOVA. The MANOVA showed that the cate gorical predictor variables— geographic location, land ownership, and credit source— had significant effects on the set of four criterion variables in this study. Further in vestigation of this multivariate relationships through step- down three-way analysis of variance showed results that partly confirmed hypothesis 4.3. These are as follows:
1. Geographic location had significant main effects on
all criterion variables except perception level. Figure 7. The interaction effect of land ownership and geographic and land ownership of effect interaction The 7. Figure
Z O H H M n j O W S location on perception level. on perception location 22 2] 23 24 GEOGRAPHIC LOCATION GEOGRAPHIC at West East A B 4 - -a- (partr-3 full- & land owners)land (non-land owners)
88 89
Western small-scale rice farmers had higher agricul
tural knowledge adoption level, average rice yield,
and participation level in the activities of their
Samahang Nayon organization.
2. Geographic location interacted with land ownership
producing a significant effect on perception level.
3. Land ownership interacted with credit source produc
ing significant effects on all criterion variables
except agricultural knowledge adoption level.
Further investigation of the two-way interaction by plotting cell means and testing for significant differences of the least square means revealed patterns of relationships that are difficult to explain, necessitating further research on the subject. Generally, it was shown that:
1. Among non-land owners, those without credit had
higher scores on participation and perception levels
and higher average rice yields than usurer creditors.
2. Among part-owners and owners, credit source did not
make a difference on average rice yield. However, for
participation and perception levels, bank-relative
creditors scored higher than those without credit
on both participation and perception levels.
3. In the east, non-land owners had higher perception
level scores than part-owners and owners. In the
west, however, the reverse was true. 90
Results of Canonical Variate Analyses
As stated in Chapter 1, one of the objectives of this research was to determine the relationship between the set of twenty-two numeric predictor variables and the set of four criterion variables. Canonical variate analysis accom plishes this objective (see Darlington, Weinberg, & Walberg,
1973). To understand better the relationships between the two sets of variables and to increase the utilization of the findings, a canonical variate analysis was performed for each group of respondents according to geographic location.
To facilitate the understanding of canonical variates and the accompanying multivariate structures of relationships and, thus, enhance substantive interpretation, step-down analyses (suggested by Roy, 1958) were also performed. That is, a reduced canonical variate analysis between each subset of the predictor variables (demographic, technical, physical, and political) and the set of four criterion variables was performed. Tatsuoka (1971), however, warned on the difficul ty of interpreting canonical analysis:
With real data, one would seldom expect . . . clear-cut (and stereotype-confirming) results. But this does not detract from the potential value of canonical analysis. It would simply mean that the dimensions of one domain . . . that are strongly associated with those of another domain . . . are not necessarily susceptible to "meaningful" verbal description within the framework of our intuitive, everyday concepts. It may be that subsequent research will show that precisely these "nonintuitive" dimensions represented by the canonical var iates are of greater scientific import (p. 191). 91
The analyses generated four important statistics: Cano nical correlations, eigenvalues, structure coefficients, and redundancies. As stated in Chapter 3, canonical correlations represent the product moment correlation between two variates and the squared canonical correlations (eigenvalues) repre
senting the "proportion of variance of each variate that is predictable from the variate of the other set" Weis, 1975, p. 334). Redundancy is a measure of the predictive power of a canonical variate. Specifically, it is the "proportion of variance extracted by the factor times the proportion of shared variance between the factor and the corresponding canonical factor of the other battery" (Cooley & Lohnes,
1971, p. 170). Redundancy, therefore, answers this question:
How much variance of one variable set is predictable from the other variable set?
For each significant canonical correlation, structure coefficients for each of the original variables were obtain ed to understand the nature of the linear relationship bet ween the sets of variables that produced the particular ca nonical correlation. They tend to be most useful relative to
the interpretation of canonical factors when a variable set
is characterized by multicolinearity (Levine, 1977). Struc
ture coefficients describe the strength of the relationship
between the original variables and the canonical variate,
and they are interpreted as if they are factor loadings
(Colley & Lohnes, 1971). To interpret the resultant canonical 92 factors, greatest emphasis was placed on the sign and magni tudes of the structure coefficients. Additionally, those with structure coefficients that are lower than -.25, though shown in appropriate tables, are suppressed or eliminated in the discussion. This step is necessary (1) as a means of highlighting those variables with moderate to strong load ings in the structure of relationships (see Cooley & Lohnes,
1971) and (2) as a means of simplifying the complicated multivariate relationships.
Eastern Leyte. Results of the canonical variate analy ses are summarized in Table 10. The analysis on eastern Ley te generated three significant canonical variates (]o<_.0087).
The correlation for the first canonical variate (and the re sulting dimension) was a high .834, accounting for 70 per cent of the variability between the linear combinations. The redundancy measure computed for this principal canonical variate was .20; thus, 20 percent of the total variability in the four criterion variables set can be predicted by the first linear combination of the twenty-two numeric predictor variables.
The second canonical variate had a canonical correlation of .691 with an eigenvalue of .478 and a redundancy of .13 for the criterion variable set. Canonical correlation for the third variate was .605. Eigenvalue was .366 and redun dancy was .10 for the criterion variable set. Total redun dancy for the criterion variable set was .47; thus, 47 TABLE 10
SUMMARY OF THE CANONICAL CORRELATION ANALYSES BETWEEN THE SET OF TWENTY-TWO PREDICTOR VARIABLES AND THE SET OF FOUR CRITERION VARIABLES
Variate Canonical Eigenvalue Asymptotic df Significance Criterion Variable Set Correlation r2 Chi square Level Redundancy Trace Coefficient
Eastern Leyte
1 .834 .696 221.60 88 .0001 .20 .29 2 .691 .478 119.92 63 .0001 .13 .27 3 .605 .366 64.32 40 .0087 .10 .27 4 .506 .256 25.32 19 .1502 .04 .17 Total .47 Total 1.00
Western Leyte
1 .797 .636 186.67 88 .0001 .27 .43 2 .638 .407 100.?9 63 .0020 .09 .22 3 .582 .339 55.62 40 .0513 .07 .21 4 .459 .211 20.22 19 .3815 .03 .14 Total .46 Total 1.00
UO U> 94 percent of the total variability of the four criterion varia ble set can be predicted from the four linear combinations of the twenty-two numeric predictor variables.
These results illustrate the difference between statis tically and substantively significant results (Myers, 1977) .
The potential meaningfulness of the relatively high signifi cant eigenvalues for the second and the third canonical var- iates are diminished by the relatively low redundancies. In the absence of a strong established theories on small-scale farmers, the resulting dimensions should be treated as take off points for future research.
To determine which predictor and criterion variables are most important to this multivariate relationships, the structure coefficients (factor structure) were examined.
Table 11 displays the structure coefficients associated with the canonical factors for the three significant canonical variates.
The three significant canonical correlations seemed to have identified three dimensions of small-scale rice farmers who are members of the Samahang Nayon organizations in east ern Leyte. The principal canonical variate identified a di mension which describes small-scale rice farmers who are high adoptors of agricultural technology and moderate in participation level. These farmers are likely to have Muni cipal Development Officers and Bureau of Agricultural Exten sion Technicians with good technical ability and to be well- TABLE 11 STRUCTURE COEFFICIENTS FROM THE CANONICAL ANALYSIS OF THE SET OF TWENTY-TWO NUMERIC PREDICTOR VARIABLES AND THE SET OF FOUR CRITERION VARIABLES (EASTERN LEYTE)
Variables Variate 1 Variate 2 Variate 3 C R=.834 C R=.691 C R=.605 Criterion Set 1. Agricultural Knowledge .973 -.156 .139 Adoption Level 2. Average Rice Yield .202 -.419 .810 3. Participation Level .353 .475 .486 4. Perception Level .217 .846 .370 Predictor Set 1. Age .054 -.031 .036 2. Education .020 .137 .161 3. Household Size .188 .204 .042 4. Years in Fanning .096 -.017 .203 5. Years as SN Member .070 .163 .141 6. Per Capita Monthly Income .101 -.116 .221 7. Per Capita Rice Area .182 .237 .402 8. Technical Ability, MDO .460 .034 .015 9. Frequency of Meeting, .239 .301 .093 MDO and Farmers 10. Communication Flow, .183 -.070 .006 MDO and Farmers 11. Planning Involvement .373 .354 .089 12. Participation Involvement -.010 .445 .167 13. Technical Ability, BAEX .515 -.035 .074 14. Frequency of Meeting, .129 .264 .104 BAEX and Farmers 15. Support of Village Officials -.069 .118 .399 16. Support of Town Officials .385 .031 .059 17. Soil Fertility .380 .453 .295 18. Irrigation Sufficiency .308 .464 .008 19. Road Condition .362 .409 .006 20. Flood Control Adequacy -.056 .438 .410 21. Distance, Farm to Town Market .055 .083 .199 22. Distance, Farm to All- -.706 .078 .253 Weather Road 96 involved in the planning of their Samahang Nayon organiza tions that are well supported by town officials. In addition, their farms are likely to be fertile, have good irrigation, and close proximity to good farm-to-market roads. This di mension is illustrated as follows:
First Dimension (Redundancy = 20%)
Predictor Variable Criterion Variable Set Set
Moderate Loadings Moderate Loadings
Technical Ability, MDO (+) Participation Level (+) Planning Involvement (+) Technical Ability, BAEX (+) High Loadings Support of Town Offi cials (+) Agricultural Knowledge Soil Fertility (+) Adoption Level (+) Road Condition (+) . .
High Loadings
Distance, Farm to All- Weather Road (-)
The first dimension of small-scale farmers identified by the first significant canonical correlation seem to sup port some time-honored principles in the diffusion of inno vations: knowledge acquisition and performance of extension clientele are influenced by the quality of extension agents,
support of the community, and physical factors. Interesting
ly, demographic variables did not appear as important
factors.
The second dimension presents a profile of farmers who
have low average rice yield; nonetheless, these farmers have moderate participation in the activities of the Samahang 97
Nayon and are highly optimistic about the future of their organization. These farmers are likely to: (1) have infertile farms with poor irrigation and flood control; (2) meet more often with their Municipal Development Officers and Bureau of Agricultural Extension Technicians; and (3) be well in volved during the planning stage of their Samahang Nayon organizations. For better understanding of the second dimen sion, a matrix of the multivariate relationships is pres ented :
Second Dimension (Redundancy = 13%)
Predictor Variable Criterion Variable Set Set
Moderate Loadings Moderate Loadings
Frequency of Meetings, Average Rice Yield (-) MDO and Farmers (+) Participation Level (+) Planning Involvement (+) Participation Involve- High Loading ment (+) Frequency of Meetings, Perception Level (+) BAEX and Farmers (+) Soil Fertility (-) Irrigation Sufficiency (-) Road Condition (-) Flood Control Adequacy (-)
The second dimension indicates the importance of:
(1) technical factors in explaining participation and per ception levels and (2) physical factors as limiting factors
in the average rice yields of small-scale farmers. The pro
file, however, provides an interesting phenomenon: a group of small-scale rice farmers who, in spite of low production,
are moderately participating in the organizational activities of the Samahang Nayon and are highly optimistic about the future of their organization. This seems to defy the logic of the culture of poverty thesis— that farmers with low pro ductivity are bound to be pessimistic and anti-group; thus, small-scale farmers with very low rice yields are expected to have low participation and perception levels. One possible explanation for this interesting contrast is that this group of farmers have other major sources of income that were not fully captured by the instrument of this study. In effect, their positive participation and perception are, in fact, product of other positive income sources.
The third significant canonical variate identifies a group of small-scale rice farmers who have very high average rice yield, good participation in the activities of the Sa mahang Nayon organization, and moderately optimistic about the future of their organization. These farmers are likely to have very small rice farms with poor soil and flood con
trol and with close proximity to all-weather road. Addition ally, they are members of Samahang Nayon organizations that
enjoy good support from village officials.
The matrix presented below illustrates the relationship between the predictor and criterion variable sets in the
third significant canonical variate. 99
Third Dimension (Redundancy = 10%)
Predictor Variable Criterion Variable Set Set
Moderate Loadings Moderate Loadings
Per Capita Rice Area (-) Participation Level (+) Support of Village Perception Level (+) Officials (+) Soil Fertility (-) High Loading Flood Control Adequacy (-) Distance of Farm to Average Rice Yield (+) All-Weather Road (-)
The third canonical correlation identifies a very inter esting profile of eastern Leyte small-scale farmers— farmers who have high average rice yield in spite of very small per capita rice area and negative (harsh) physical factors of soil fertility and flood control. One possible explanation is that these farmers might have used indigenous technology in in farming such as the use of farm and household refuse as fertilizers. These practices were not included in determin ing the agricultural knowledge adoption level. This finding seems to support the main thesis of small-scale farmer deve lopment (Biggs, 1974; World Bank, 1974; Food and Agriculture
Organization, 1975) . That is, small-scale farmers are pro ductive because they are efficient users of available re sources .
The three significant dimensions of small-scale rice farmers presented a multi-faceted empirical view of the re lationships between the predictor and criterion variable set.
To simplify these complex multivariate relationships, reduced canonical variate analyses were conducted. The structure 100
coefficients and redundancies of the first canonical variates are summarized in Table 10.
Of the twenty-two variables, twelve showed moderate
(-.25 to -.64 structure coefficients) to strong (>.65) rela tionships. The computed redundancies showed that for eastern
Leyte, the order of decreasing importance of the predictor variable subsets in explaining the variation of the criterion variable set were as follows: first, physical; second, tech nical; third, demographic; and fourth, political factors.
A graphic presentation of the multivariate relation
ships between the set of predictor variables and the set of criterion variables on the resultant factor loadings (struc ture coefficients) from the reduced canonical analyses of the data from eastern Leyte is shown in Figure 8. The num bers in every criterion variable correspond to the number
for the predictor variable subsets, and the letters and
signs in parentheses indicate the direction and magnitude
of the relationships. The listed variables for every subset
are those that show moderate to strong relationships. Thus,
it can be seen that agricultural knowledge adoption level was moderately affected by demographic variables and highly
affected by technical, political, and physical factors.
Average rice yield was highly affected by demographic varia
bles and moderately affected by physical factors. For the
third dependent variable, participation level, technical and TABLE 12 STRUCTURE COEFFICIENTS FROM THE CANONICAL ANALYSES OF EACH SUBSET OF THE NUMERIC PREDICTOR VARIABLES AND THE SET OF CRITERION VARIABLES (EASTERN LEYTE)
Predictor Structure Criterion Structure Variables______Coefficients Variables Coefficients Demographic Subset
Variate 1: Canonical R=.530 Eigenvalue=:. 280 Chi square=42.82 Sig. level= .036 1. Age .088 1. Agricultural .557 2. Education -.017 • Knowledge 3. Household Size -.034 2. Average Rice .966 4. Years in Farming .223 Yield 5. Years as SN -.224 3. Participation .032 Member Level 6. Per Capita .326 4. Perception -.040 Monthly Income Level 7. Per Capita Rice -.647 Area Redundancy for the set of criterion variables given demographic variables .09 rechnical Subset Variate 1: Canonical R=.640 Eigenvalues 409 Chi square=72.74 Sig. levels 00001 1. Technical Ability .558 1. Agricultural .814 MDO Knowledge 2. Frequency of .423 2. Average Rice -.003 Meeting, MDO Yield 3. Communication .184 3. Participation .518 Flow, MDO & Level Farmers 4. Perception .542 4. Planning .602 Level Involvement 5. Participation .192 Involvement 6. Technical Ability .592 BAEX 7. Frequency, Meeting .273 BAEX & Farmers Redundancy for the set of criterion variables given technical variables .13 TABLE 12 (continued)
Predictor Structure Criterion Structure Variables______Coefficients_____Variables Coefficients Political Subset Variate 1: Canonical Rs424 Eigenvalues 180 Chi square=25.0 Sig. l e v e l s 002 1. Support of -.246 1. Agricultural .824 Village Officials Knowledge 2. Support of .897 2. Average Rice .234 Town Officials Yield 3. Participation -.176 Level 4. Perception .282 Level Redundancy for the set of criterion variables given political variables .04 Physical Subset Variate 1: Canonical Rs744 Eigenvalue= .544 Chi square=115.25 Sig. level= .0001 1. Soil Fertility .506 1. Agricultural .962 2. Irrigation .440 Knowledge Sufficiency 2. Average Rice .403 3. Road Condition .515 Yield 4. Flood Control .044 3. Participation .336 Adequacy Level 5. Distance of Farm .116 4. Perception -.045 to Town Market Level 6. Distance of Farm -.813 to All-Weather Road Redundancy for the set of criterion variables given physical variables Predictor Variable Set Criterion Variable Set
1. Denpgrahic
{+ M) Per Capita Monthly income -*(+ M) (- H) Per Capita Rice Area --- (+ H) 2. Technical Agricultural Knowledge ■4 (+ H) — Adoption Level (+ M) 'Technical Ability, MDO - (+ M) Frequency of Meetings, -* (+ H) MDO & Farmers------(+ M) Planning Involvement — (+ M) Technical Ability, BAEX (+ M) Frequency of Meetings, (+ H) BAEX & Farmers------Average Rice Yield ■* (+ M) 3. Political (+ H) Support of Town Officials -- 4- — (+ M) — 1 2 4. Physical Kurticipation Level (+ M) M (+ M) 9oil Fertility ------(+ M) Irrigation Sufficiency (+ M) Road Condition (- H) Distance of farm to ->(+ M) All-Weather Road — > Perception Level ->(+ M)
Where: - or + indicate direction of M - moderate relationship H - strong relationship
Figure 8. Graphic presentation of the multivariate relationships between the set of predictor variables and the set of criterion variables on the resultant structure coefficients from the reduced canonical analyses on eastern Leyte. 104 physical factors had moderate effects. For perception level, both technical and political factors had moderate effects.
Western Leyte. The canonical variate analysis for west ern Leyte respondents had three significant (£ < .0517) canonical correlations (Table 10). The principal canonical factor was .797 with an eigenvalue of .636; thus, 64 percent of the variation in the first canonical variate can be ex plained by the variability between the two first linear com binations. Redundancy, however, was .27. This indicated that only 27 percent of the total variance in the four-variable criterion set can be predicted by the first canonical factors of the twenty-two numeric predictor variable set.
The second canonical variate had a canonical correla tion of .638 with an eigenvalue of .407. Redundancy was com puted to be .09. Canonical correlation for the third variate was .582. Eigenvalue was .339 and redundancy, .07. It should be pointed out that the potential meaningfulness of the rela tively high significant eigenvalues for the second and third canonical variates are diminished by the relatively low redundancy. Even with the second and third linear combina tions of the predictor variable set, only 16 percent of the of the four-variable criterion set can be explained.
To determine which predictor variable and criterion variable are most important to this multivariate relation ship, the structure coefficients (factor structure) were examined. Table 13 displays the structure coefficients TABLE 13 STRUCTURE COEFFICIENTS FROM THE CANONICAL ANALYSIS OF THE SET OF TWENTY-TWO NUMERIC PREDICTOR VARIABLES AND THE SET OF FOUR CRITERION VARIABLES (WESTERN LEYTE)
Variables Variate 1 Variate 2 Variate 3 ______C R=.797 C R=.638 C R=.605 Criterion Set 1. Agricultural Knowledge .677 .299 .671 Adoption Level 2. Average Rice Yield .524 .657 -.352 3. Participation Level .626 -.555 -.050 4. Perception Level .753 -.247 -.454 Predictor Set
1. Age -.107 .019 .222 2. Education .538 -.010 .005 3. Household Size .060 .287 -.091 4. Years in Fanning .200 .034 .149 5.' Years as SN Member -.388 -.156 -.469 6. Per Capita Monthly Income .486 -.017 .167 7. Per Capita Rice Area .158 -.253 .478 8. Technical Ability, MDO .039 -.319 .079 9. Frequency of Meeting .343 -.055 -.042 MDO and Farmers 10. Communication Flow, .166 -.399 .040 MDO and Farmers 11. Planning Involvement .286 -.481 .339 12. Participation Involvement .636 -.399 .155 13. Technical Ability, BAEX .213 -.393 .121 14. Frequency of Meeting, .569 .025 -.004 BAEX anf Farmers 15. Support of Village Officials .270 -.341 -.426 16. Support of Town Officials .507 -.072 -.222 17. Soil Fertility .245 .090 .464 18. Irrigation Sufficiency .522 .111 .012 19. Road Condition .071 .088 .166 20. Flood Control Adequacy .159 -.026 -.083 21. Distance, Farm to .135 .111 .509 Town Market 22. Distance, Farm to -.017 .236 -.175 All-Weather Road 106 associated with the canonical factors for the first three significant canonical variates.
The structure coefficients seem to have identified three profiles of small-scale rice farmers. The first dimen sion identifies a group of farmers who are moderate in aver age rice yield and .participation level, high adoptors of agricultural knowledge, and very optimistic about the future of their Samahang Nayon organizations. These farmers are likely to have higher education, shorter period of member ship in the Samahang Nayon organization, and higher per capita monthly income. Furthermore, they are likely to have fertile ricefields with sufficient irrigation and more fre quent meetings with both the Municipal Development Officers and the Bureau of Agricultural Extension Technicians. Addi tionally, they have greater degree of participation in the implementation stage of their Samahang Nayon organizations, and their organizations enjoy good support from village and town officials. The following matrix illustrates the farmers in dimension one. 107
First Dimension (Redundancy = 27%)
Predictor Variable Criterion Variable Set Set
Moderate Loadings Moderate Loadings
Education (+) Average Rice Yield (+) Years as SN Member (-) Participation Level (+) Per Capita Monthly Income (+) High Loadings Frequency of Meetings, MDO and Farmers (+) Agricultural Knowledge Planning Involvement (+) Adoption Level (+) Participation Perception Level (+) Involvement (+) Frequency of Meetings, BAEX and Farmers (+) Support of Village Officials (+) Support of Town Officials (+) Soil Fertility (+) Irrigation Sufficiency (+)
The first dimension shows an almost "ideal" group of
small-scale rice farmers in western Leyte. It appears that
all four categories of predictor variables affected the
criterion variable set with technical factors accounting for
one-half of the variables that showed moderate relation
ships. As in the first canonical correlation in eastern
Leyte, this finding partly lends support to the strategy
expounded in agricultural extension regarding the use of
participatory involvement (clientele and community) in
extension activities. That is, the target clientele and the
community need to be involved in all phases of extension
activities starting from planning to evaluation of programs. 108
The second dimension identifies small-scale rice farmers with high average rice yield but with low participation lev el in the activities of their Samahang Nayon. These farmers are likely to have (1) Municipal Development Officers and
Bureau of Agricultural Extension Technicians with poor tech nical ability, (2) poor communication with their Municipal
Development Officers, and (3) low involvement in the plan ning stage of the Samahang Nayon organization. Additionally, these farmers are likely to be members of Samahang Nayon organizations that are not well-supported by the village officials.
For better understanding of the relationships between the predictor and criterion variable sets in the second sig nificant canonical correlation, a matrix was constructed:
Second Dimension (Redundancy = 9%)
Predictor Variable Criterion Variable Set Set
Moderate Loadings Moderate Loadings
Household Size (+) Agricultural Knowledge Per Capita Area (-) Adoption Level (+) Technical Ability, MDO (-) Participation Level (-) Communication Flow (-) Perception Level (-) Planning Involvement (-) Technical Ability, BAEX (-) High Loading Support of Village Officials (-) Average Rice Yield (+)
From the list of predictor variables, it appears that
there are no logical explanations for the high loadings
obtained in average rice yield. The low participation and 109 perception levels, however, seem logically explained by the negative loadings of the four technical factors and the sup port of village officials. This findings support the earlier evidence (first dimension) regarding the moderate positive relation between technical variables and participation level.
The third dimension identifies small-scale rice farmers who are high adoptors of agricultural knowledge, yet they have low rice yield. Additionally, these farmers have low participation level and are pessimistic about the future of their organization. These farmers are likely to (1) have big ger per capita rice area, (2) have farms that are located far from the town market, and (3) have joined organizations that are not well-supported by village officials. Addition ally, they have farms with fertile soil.
As a means of looking at the multivariate relation ships in the third significant canonical correlation, a matrix is presented below:
Third Dimension (Redundancy = 7%)
Predictor Variable Criterion Variable Set Set
Moderate Loadings Moderate Loadings
Years as SN Member (-) Average Rice Yield (-) Per Capita Rice Area (+) Perception Level (-) Planning Involvement (+) Support of Village High Loadings Officials (-) Soil fertility (+) Agricultural Knowledge Distance, Farm to Town Adoption Level (+) Market (+) 110
The foregoing structure of relationship illustrates what Tatsuoka (1971, p. 191) calls as those dimensions that are not "susceptible to 'meaningful' verbal description within the framework of our everyday, intuitive concepts."
Nonetheless, this finding shows an interesting problem that needs further investigation.
As in eastern Leyte, the three significant dimensions of small-scale rice farmers in the western part of the prov ince present a multi-faceted empirical view of the relation ships between the predictor and dependent variable sets. As a means of simplifying these complex multivariate relation ships, reduced canonical variate analyses were performed.
The structure of relationships resulting from the first canonical variates and the computed redundancies is display ed in Table 14.
Of the twenty-two numeric predictor variables, eight were moderately (-.25 to -.64 structure coefficients) affect ing the criterion variable set while six predictor varia bles were strongly affecting it (^.65). The computed redun dancies showed that for western Leyte, the most important subsets of predictor variables that explained the variabili ty of the dependent variable set were demographic and tech nical factors. These were followed by physical and political factors.
A graphic presentation of the multivariate relation ships between the set of predictor variables and the set of TABLE 14 STRUCTURE COEFFICIENTS FROM THE CANONICAL ANALYSES OF EACH SUBSET OF THE NUMERIC PREDICTOR VARIABLES AND THE SET OF CRITERION VARIABLES (WESTERN LEYTE)
Predictor Structure Criterion Structure Variables______Coefficients_____ Variables Coefficients Demographic Subset Variate 1: Canonical R=.627 Eigenvalue3.393 Chi square=77.58 Sig. level=.0001
1. Age -.134 1. Agricultural .722 2. Education .650 Knowledge 3. Household Size .170 2. Average Rice .739 4. Years in Farming -.254 Yield 5. Years as SN -.499 3. Participation .444 Member Level 6. Per Capita .568 4. Perception .614 Monthly Income Level 7. Per Capita .086 Rice Area Redundancy for the set of criterion variables given demographic variables .16 technical Subset Variate 1: Canonical R=.631 Eigenvalue3 .398 Chi square=77 .30 Sig. level3 .00001 1. Technical Ability .148 1. Agricultural .634 MDO Knowledge 2. Frequency of .423 2. Average Rice .244 Meeting, MDO Yield 3. Communication Flow , .316 3. Participation .737 MDO & Farmers Level 4. Planning .519 4. Perception .743 Involvement Level 5. Participation .899 Involvement 6. Technical Ability .382 Baex 7. Frequency, Meeting .671 BAEX & Farmers Redundancy for the set of criterion variables given technical variables .15 TABLE 14 (continued)
Predictor Structure Criterion Structure Variables Coefficients Variables Coefficients
Political Subset Variate 1: Canonical R=.506 Eigenvalues256 Chi square=33.86 Sig. levelsOOOl 1. Support of .704 1. Agricultural .157 Village Officials Knowledge 2. Support of .834 2. Average Rice .332 Town Officials Yield 3. Participation .630 Level 4. Perception .961 Level Redundancy for the set of criterion variables given political variables .09 Physical Subset Variate 1: Canonical R=.551 Eigenvalue=.304 Chi square=58.18 Sig. levels0001 1. Soil Fertility .557 1. Agricultural .964 2. Irrigation .690 Knowledge Sufficiency 2. Average Rice .335 3. Road Condition .213 Yield 4. Flood Control .158 3. Participation .339 Adequacy Level 5. Distance of Farm .458 4. Perception .391 to Town Market Level 6. Distance of Farm -.058 to All-Weather Road Redundancy for the set of criterion variables given physical variables .10 113 criterion variables on the resultant factor loadings (struc ture coefficients) from the reduced canonical analyses of the data from western Leyte is displayed in Figure 9. It is shown that for agricultural knowledge adoption level, demo graphic and physical factors showed strong effects while technical factors showed moderate effects. For average rice yield, demographic factors had strong effects while politi cal and physical factors exhibited moderate effects.
All four predictor variable subsets showed moderate to strong effects on participation and perception levels. Demo graphic and physical factors exhibited strong effects.
Political variables, however, affected both variables in different ways. It had moderate effects on participation level, but it had strong effects on perception level.
Summary of canonical variate analyses. The canonical variate analysis was used to test the relationship between the set of twenty-two numeric predictor variables and the set of four criterion variables. Five separate canonical analyses were performed for each geographic location or a total of ten analyses for the whole study. The results show ed that for each geographic location, three significant canonical correlations were identified. The high eigenvalues for the significant canonical correlations, however, were diminished by the relatively small redundancies. Examination of the resulting structure of relationships for every Predictor Variable Set Criterion Variable Set
1. Demographic
(+ H) Education ------> (+ H) — 1 (- M) Years in Fanning------(+ M) Per Capita Monthly Income -4 (+ M) — 2 Agricultural Knowledge (- M) Years as SN Member----- Adoption Level 2. Technical (+ H) 4 (+ M) Frequency of Meetings, MX) & Farmers (+ M) Ccaimmication Involvement — -*(+ H) 1 (+ M) Planning Involvement----- (+ H) Participation Involvement — -»(+ M) 3- Average Rice Yield (+ M) Techical Ability, BAEX---- (+ H) Frequency of Meetings, -* (+ M) 4 BAEX & Farmers ------3. Political -*(+ M) n 1 {+ H) Support of Village Officials — r (+ H) Support of Town Officials --- •*" (+ H) 2 Participation Level 4. Physical -* (+ M) —
(+ M) Soil Fertility ------* (+ M) (+ H) Irrigation Sufficiency — (+ M) Distance of Farm to All- Weather Road------> (+ M) —
Where: - or + indicates the direction -H+ H) — of relationships -v Perception Level M - moderate relationship (+ H) — 3 H - strong relationship -* (+ M ) 14
Figure 9. Graphic presentation of the multivariate relationships between the set of predictor variables and the set of criterion variables on the resultant structure coefficients from the reduced canonical analyses on western Leyte. 115 significant canonical correlation identified three dimensions of small-scale rice farmers from every geographic location.
The findings in the canonical variate analyses illus trated the complex multivariate relationships between the predictor and criterion variable sets in this study. None theless, there are patterns of relationships that are con sistent in both geographic locations. The differences and similarities in the multivariate relationships between the two sets of variables are illustrated in a matrix (see
Figure 10) derived from the reduced canonical analyses and the resulting structures of relationship. In view of the amount of data simplification involved to arrive at the matrix, there is a need for caution in arriving at generali zations .
The predictor variable subsets exhibited varying strengths of effects on the criterion variables under dif ferent geographic locations. It is beyond the scope of this research to explain this important observation that deserves further research. Obviously, the results of which have impli cations both for program improvement and policy making. Predictor Variable Criterion Variable Set Subsets Agricultural Average Participation Perception Knowledge Rice Yield Level Level Adoption
East West East West East West East Wesi
Demographic +M +H +H +H A +M * +M Technical +H +M ** +M +H +M +H Political +H ** +M * +M +M +H Physical +H +H +M +M +M +M * +M
where: * = weak effect +M = positive moderate effect +H = positive strong effect
Figure 10. Matrix showing the comparative effects of the different subsets of predictor variables on the criterion variable set by geographic location. CHAPTER V
SUMMARY, CONCLUSIONS, AND IMPLICATIONS
Summary
The Problem
This study examined the relationships between the sets of selected demographic and situational variables and the set of four outcome measures as a result of participating in the rural development program of Samahang Nayon by small- scale rice farmers in Leyte, Philippines.
The predictor variables in this study were examined, classified, and defined. These variables were divided into two sets: categorical, composed of three variables and nu meric, composed of twenty-two variables. The numeric varia bles were further categorized into the following: demograph ic (seven variables), technical (seven variables), politi cal (two variables) , and physical (six variables) factors.
The criterion variables were composed of four outcome measures as a result of the participation of small-scale rice farmers from Leyte in the Philippine rural development program of Samahang Nayon. These outcome variables were: agricultural knowledge adoption level in rice production, average rice yield in cavans per hectare, participation level in the activities of the Samahang Nayon organization,
1 1 7 118 and perception level about the future of the Samahang Nayon organization.
Methodology
Through multi-stage sampling procedure, a random sample of 100 small-scale rice farmers, who were members of the
Samahang Nayon organization, was drawn from each of the eastern and western parts of the province of Leyte, Philip pines. These 200 randomly selected small-scale rice farmers constituted the respondents of this study.
The major aim of the methodology used was to collect complete and accurate data from the respondents. To achieve this objective, an interview schedule was developed and used to collect data from the respondents. Data collection was performed by hired interviewers under the supervision of the
Extension Research and Development Division staff of the
Visayas State College of Agriculture in Leyte, Philippines.
Preliminary runs were made to detect errors in coding, keypunching, and entries with the use of the Statistical
Analysis System (SAS, 1979) computer program. After errors were corrected, data were analyzed with the use of SAS and, to some extent, BMDP (Biomedical Statistical Package, 1977) computer programs.
Descriptive statistics such as frequencies, percent
ages, means, and standard deviations were computed to sum marize the data regarding the characteristics of the small- 119
scale rice farmers. Additionally, zero-order correlations were calculated among numeric variables to measure the asso ciation between variables.
Inferential statistics were generated to test the hypo theses of the study. The specific inferential statistics used were as follows:
1. Chi square was used to test whether there was signi
ficant differences between the eastern and western
respondents in the distribution of the following
nominal variables: status of land ownership and
credit source.
2. Multivariate analysis of variance (MANOVA) was used
to test if:
o Significant difference existed between the
eastern and western respondents in the twenty-
two numeric predictor and four criterion
variables.
o Significant relationship existed between the
set of categorical variables (credit source,
land ownership, and geographic location) and
the set of four criterion variables. The
second MANOVA was, therefore, used to test
research hypothesis 4.3 of the study.
3. Canonical variate analysis was used to test research
hypotheses 4.1 and 4.2. That is, it was employed to
determine the kinds of significant relationships 120
that may exist between the set of twenty-two numeric
predictor variables and the set of four continuous
criterion variables.
Findings
The chi square test indicated a significant difference
(£ 4 .005) between the two groups of respondents in terms of
land ownership and credit source.
The MANOVA showed that eastern and western Leyte re
spondents were significantly different (p = .0001) from each other. Follow-up multicomparison procedures showed that this was attributed to nine numeric predictor variables and two criterion variables. A summary profile showing the similari ties and differences between the two groups of respondents
is presented in Table 15.
A three-factor multivariate analysis of variance
(MANOVA) showed a significant main effect of geographic
location and significant two-way interaction effects of
(1) geographic location and land ownership and (2) land ownership and credit source on the set of four criterion variables. Further investigation of these significant effects was made with the use of a step-down three-way univariate
analysis of variance. The following results confirmed in
part research hypothesis 4.3 of the study:
1. Geographic location significantly affected agricul
tural knowledge adoption level, average rice yield, 121
TABLE 15 A SUMMARY PROFILE OF EASTERN AND WESTERN LEYTE FARMERS Means Variables East West Demographic
Age (years) 48.12 48.20 Education (years) 4.98 3.82 Household Size 6.65 5.89 Years in Farming 24.37 28.00 Years as SN Member 6.24 6.23 Per Capita Monthly Income (P) 100.98 136.13 Per Capita Rice Area(Ha/person) .18 .24 * Land Ownership 49% 72% (non-owners) (non-owners) Technical
* Technical Ability, MDO‘ 3.25 2.88 (slightly above (slightly below average) average) Frequency of Meeting, MDO and Farmers (April-June 1980) .13 ,13
Communication Flow, MDO and 3.25 2.88 Farmers3 (slightly above (slightly below average) average) Planning Involvement3 3.85 2.49 (good) (slightly below average) Participation Involvement3 1.83 2.43 (poor) (slightly below average) Technical Ability, BAEX3 3.72 3.17 (good) (average) Frequency, Meeting of BAEX 1.67 2.65 and Farmers (April-June 19 80) Political
* Support of Village Officials3 2.85 3.78 (slightly below average) (good) Support of Town Officials3 3.04 3.06 (average) (average) * Credit Source 28% none 43% none 47% usurer 25% usurer 122
TABLE 15 (CONTINUED)
Means Variables East West Physical * Soil Fertility13 3.18 2.83 b (good) (almost good) * Irrigation Sufficiency 1.80 2.28 K (poor) (poor) Road Control Adequacy 2.41 2.42 h (poor) (poor) Flood Control Adequacy 2.62 2.44- (almost good) (poor) * Distance of Farm to Town Market (kilometers) 3.62 8.32 * Distance of Farm to All- Weather Road (kilometers) .63 1.47 Criterion * Agricultural Knowledge Adoption Level (average % adoption) 35.36 54.00 * Average Rice Yield a(Cav/Ha) 47.85 62.86 Participation Level 3.64 3.84 a (good) (good) Perception Level 3.12 3.15 (average) (average)
*Significant difference at £ < .02
aRated by respondents in a five-point Likert-type scale with ^ = very poor, 2 = poor, 3 = neither good nor poor, 4 = good, and 5 = very good. bRated by appropriate government technician in a four-point Likert-type scale with 1 = very poor, 2 = poor, 3 = good, and 4 = very good. 123
and participation level of small-scale rice farmers.
Compared with eastern respondents, western small-
scale rice farmers had higher agricultural knowledge
adoption level, average rice yield, and participa
tion level.
2. Geographic location interacted disordinally with
credit source. It had significant effect on percep
tion level. In the east, non-land owners had higher
perception scores than part-owners and owners. In
the west, the reverse was true.
3. Land ownership and credit source had two-way disor-
dinal interaction effects on all criterion variables
except agricultural knowledge adoption level. Among
non-land owners, those without credit had higher
average rice yield, participation level, and percep
tion level than usurer creditors. Among part-owners
and owners, credit source did not make a difference
on average rice yield. However, for participation
and perception levels, bank-relative creditors
scored higher than those without credit on both
participation and perception levels.
These findings partly supported the major thesis ex pounded by conflict theorists, i.e., socio-structural varia bles are major determinants of the performance and outlook of peasants. It should be noted, however, that the effects of socio-structural variables in this study were not simple 124 main effects. They interacted with each other and with geog raphic location— an important variable to consensus theo rists— thus produced complicated relationships.
A canonical variate analysis was performed for each group of respondents according to geographic location for better understanding of the multivariate relationships and for increasing the utility of the information generated. As mentioned in Chapter 4, substantive interpretation of the structure of relationships resulting from the canonical analysis may be difficult; they may not be amenable to mean ingful description. Nonetheless, they serve as "take-off
9 points" for future research.
In eastern Leyte, canonical variate analysis of the set of twenty-two numeric predictor variables and the set of four criterion variables showed three significant canonical 9 correlations with eigenvalues (R ) ranging from a high of
.696 to a moderate .366. The eigenvalues— the amount of variance in one canonical variate that is accounted for by the other canonical variate— are, however, diminished by the relatively low redundancies— the amount of variation in one variable set that can be predicted from the other set. These eigenvalues ranged from 20 percent for the first significant variate to 10 percent for the third significant variate.
Examination of the resulting structure of relationships
for every significant canonical correlation seemed to have
identified three dimensions of eastern Leyte small-scale 125 rice farmers who are participating in the rural development program of the Samahang Nayon. The resulting structure of relationships that showed the multivariate relationships between the predictor variable set and the criterion varia ble set were difficult and complicated to interpret. As a method of simplification, reduced canonical variate analyses were conducted. It was shown that of the twenty-two numeric predictor variables, the following twelve showed strong to moderate relationships with the set of four criterion variables.
Predictor Variable Set
Moderate Positive Loadings
Per Capita Monthly Income Technical Ability, Municipal Development Officers Frequency of Meetings, Municipal Development Officers and Farmers Planning Involvement Technical Ability, Bureau of Agricultural Extension Technicians Frequency of Meetings, Bureau of Agricultural Extension Technicians and Farmers Soil Fertility Irrigation Sufficiency Road Condition
High Positive Loading
Support of Town Officials
High Negative Loadings
Per Capita Rice Area Distance of Farm to All-Weather Road
The foregoing list of predictor variables indicates the relative importance of the technical and physical factors in explaining the variability of the criterion variable set. 126
The computed redundancies confirmed this when they showed the relative importance of the different subsets of predic tor variables in explaining the variability of the criterion variable set: first, physical variables; second, technical variables; third, demographic; and fourth, political varia bles.
In terms of the relative specific effects of the dif ferent subsets of predictor variables in each of the varia bles in the criterion set, the reduced canonical analyses show the following results:
Criterion Variable Set
Agricultural Knowledge Adoption Level
Demographic, moderately positive Technical, strongly positive Political, strongly positive Physical, strongly positive
Average Rice Yield
Demographic, strongly positive Physical, moderately positive
Participation Level
Technical, moderately positive Physical, moderately positive
Perception Level
Technical, moderately positive Political, moderately positive
In western Leyte, canonical variate analysis of the set of twenty-two numeric predictor variables and the set of four criterion variables showed three significant canonical 2 correlations with eigenvalues (R ) ranging from a high of 127
.636 to a moderate .339. The corresponding redundancies, however, of the criterion variable set were more modest.
They ranged from 27 percent for the first significant vari ate to 7 percent for the third significant variate.
Examination of the resulting structure of relationships for every significant canonical correlation seemed to have identified three dimensions of small-scale rice farmers in western Leyte who are participating in the rural develop ment program of the Samahang Nayon. As a means of sim;_ ' , ing the complicated multivariate relationships, reduc„K.i canonical variate analyses were performed. Tho pr^ro's id_ tified that the following fifteen variables showed moderate to strong relationships with the set of four criterion variables.
Predictor Variable Set
Moderate Negative Loadings
Years in Farming Years as Samahang Nayon Member
Moderate Positive Loadings
Per Capita Monthly Income Frequency of Meetings, Municipal Development Officers Communication Flow, Municipal Development Officers and Farmers Planning Involvement Technical Ability, Bureau of Agricultural Extension Technicians Soil Fertility Distance of Farm to Town Market 128
High Positive Loadings
Education Participation Involvement Frequency of Meetings, Bureau of Agricultural Extension Technicians and Farmers Support of Village Officials Support of Town Officials Irrigation Sufficiency
From the list it appears that demographic and technical factors are the most important variables that affect the four criterion variables of this study. The redundancy for each subset of predictor variables supported this observa tion— demographic and technical factors are most important
in accounting for the variation in the set of criterion variables; these are followed by physical and political
factors, respectively.
In terms of the relative effects of the different sub
sets of predictor variables in each variable in the criteri on set, the reduced canonical analyses show the following
results:
Criterion Variable Set
Agricultural Knowledge Adoption Level
Demographic, strongly positive Technical, moderately positive Physical, moderately positive
Average Rice Yield
Demographic, strongly positive Political, moderately positive Physical, moderately positive 129
Participation Level
Demographic, moderately positive Technical, strongly positive Political, moderately positive Physical, moderately positive
Perception Level
Demographic, moderately positive Technical, strongly positive Political, strongly positive Physical, moderately positive
The results of the canonical analyses supported research hypotheses 4.1 and 4.2 in general. A comparison of the rela tive effects of the twenty-two numeric predictor variables on the set of four criterion variables by geographic loca tion is presented in Table 16. Out of the total number of numeric predictor variables, nine variables were either moderately or strongly influencing the set of four criterion variables in both geographic locations. Additionally, nine variables had differential effects on the criterion varia ble set, i.e., each of these variables had either a moderate or strong effect only in one respondent group. These varia bles are broken down as follows: five, demographic; seven,
technical; two, political; and four, physical.
The findings of the canonical variate analyses, like
those of the studies cited in Chapter 2, show that the extreme positions expounded by either conflict or consensus
theorists do not fully explain the performance and percep
tion of small-scale rice farmers. Some variables that are
important to both perspectives were found to explain part 130
TABLE 16
COMPARATIVE EFFECTS OF THE TWENTY-TWO
NUMERIC PREDICTOR VARIABLES BY GEOGRAPHIC LOCATION
Prediction Support Research Variables of Rela Results Hypotheses 4.1 tionship East West and 4.2?
Demographic Age - * * No Education + * +H Conditional Household Size - * * No Years in Farming - * -M Conditional Years as SN Member + * -M Partly rejected Per Capita Monthly Income + +M +M Yes Per Capita Rice Area - -H * Conditional
Technical Technical Ability, MDO + +M * Conditional Frequency of Meetings, MDO and Farmers + +M +M Yes Communication Flow + * +M Conditional Planning Involvement + +M +M Yes Technical Ability, BAEX + +M +M Yes Frequency of Meetings, BAEX and Farmers + +M +H Yes Participation Involvement + * +H Conditional
Political Support of Village Officials + * +H Conditional Support of Town Officials + +H +H Yes
Physical Soil Fertility + +M +M Yes Irrigation Sufficiency + +M +H Yes Road Condition + +M * Conditional Flood Control Adequacy + * * No Distance, Farm to Market - ** No Distance, Farm to All- Weather Road + -H +H ?
Where: + or - indicates direction of relationship M = Moderate relationship H = Strong relationship * = Weak relationship 131 of the variance exhibited by the set of criterion variables
in this study.
Conclusions and Implications
The complex multivariate relationships identified in
this study confirm an obvious fact, though oftentimes for gotten in policy decisions. There are no simple and easy answers to small-scale rice farmers' development.
The results of the MANOVA raise interesting questions
that require answers beyond the scope and methodology of
this study. The two-way interaction effects of land owner
ship and credit source on average rice yield, participation
level, and perception level and the two-way interaction
effect of land ownership and geographic location on percep
tion level require answers that have significant effect on policy program improvement.
The canonical variate analyses show that demographic,
technical, political, and physical factors have multivariate
effects on agricultural knowledge adoption level, average
rice yield, participation level, and perception level that
are bound to change in pattern due to cultural differences.
Thus, changes are also expected for the relative importance
of the different subsets of predictor variables in explain
ing the variation in the criterion variable set in differ
ent cultural settings. 132
Despite methodological limitations, the study illus
trates the complexity in explaining the outcomes of small-
scale rice farmers' participation in the Philippine rural
development program of Samahang Nayon. The extreme positions
posited by both conflict and consensus theorists do not seem
to hold. The study shows a multi-faceted reality in rural
development, negating the single theory approach to formula
tion of strategies designed for small-scale rice farmers' development. Under varying conditions cf demographic, tech
nical, political, and physical factors, small-scale rice
farmers display wide variations in the four outcome varia bles, thus requiring sensitivity and flexibility and at the
same time high technical expertise on the part of the rural
development planners to design programs that are effective.
The major implication to be drawn from this study is
concerned primarily with the establishment of a basis for more comprehensive empirical research exploring the demo
graphic, technical, political, and physical variables rela
tive to outcomes of participating in small-farmer develop ment program. The complex multivariate relationships found
in this study, some of which seem to be unsusceptible to meaningful explanation, can be fully understood only with
further empirical investigation.
A replication of this study using a larger sample and
various groups of small-scale farmers in different cultural
settings would provide further empirical basis for 133 understanding the target clientele of Third World rural development programs and, at the same time, provide better scientific basis for designing extension programs. In short, there is a need for comparative research in studying differ ent commodity (i.e., groups of small-scale farmers according to crops grown) and cultural groups of small-scale farmers.
Additionally, further research is needed in identifying additional political variables and understanding the rela tionships of these factors to specific outcome measures in small-scale farmer development programs. This may have a substantial effect in increasing the redundancy of the pre dictor variable set to explain the variability in the crite rion variable set. Further, there is a need to use other research techniques such as qualitative methodologies and ethnographic techniques in understanding better the rela tionship among demographic, technical, political, and phys ical factors and the criterion variables in this study. REFERENCES
Adams, D. W. & Coward, E. W., Jr. Small-farmer Development Strategies. New York: ADC, 1972.
Agricultural Staff. Philippine Agriculture in the Last Twenty Years. Manila, Philippines: NEDA, 1975.
Alvarez-Quintero. "Evaluation of Rural Development Programs in Developing Countries: An Appraisal of Supervised Credit Programs." Ph.D. Dissertation. The Ohio State University, 1975.
Anderson, C. H. The Political Economy of Social Class. Englewood Cliffs, NJ: Prentice-Hall, Inc., 1974.
Aquino, B. "Mr. Marcos and the Future of the Philippines." The Wall Street Journal, undated.
Aziz, S. Rural Development: the Chinese Approach. In FAO, Training for Agriculture and Rural Development. Rome: FAO, 1975.
Bauer, R. (ed.) Social Indicators. Cambridge, MA: The M.I.T. Press, 1966.
Biggs, H. H. "New Perspective on Development Strategies." In H. H. Biggs & R. L. Tinnermeier (eds.), Small Farm Agricultural Development Problems.Fort Collins, CO: Colorado State University, 1974.
Bingswanger, H. P. et al. Induced Innovation: Technology, Institutions, and Development. Baltimore, MD: The John Hopkins University Press, 1978.
Blase, M. G. & Goodwin, J. B. (eds.) Readings in Inter national Agriculture Economic Development. New York: MSS Ed. Pub. Co., Inc., 197 0.
Boguslaw, R. & Vickers, G. Prologue to Sociology. Santa Monica, CA: Goodyear Publishing Co., Inc., 1977.
Castillo, G. T. All in A Grain of Rice. Los Banos: SEARCA, 1975.
134 135
______Beyond Manila; Philippine Rural Problems in Perspective. Vol. I, II, III. Laguna, Philippines: UPLB, 1977.
______"Comment: A Critical View of a Subculture of Peasantry." In Subsistence Agriculture and Economic Development. Edited by C. R. Wharton. Chicago: Aldine Publishing Co., 1969.
Chambliss, W. J. (ed.) Sociological Readings in the Conflict Perspective. Reading, MA: Addison-Wesley, 1973.
Chayanov, A. V. The Theory of Peasant Economy. Illinois: Irwin, Inc., 1966.
Clark, R. Land Reform and Peasant Market Participation on the Northern Highlands of Bolivia. LTC Reprint No. 42. Madison, WI : Land Tenure Institute, 1968.
Comptroller General of the United States. "Report to the Congress." Disincentives to Agricultural Production in Developing Countries. Dept, of State and Other Agencies, Nov. 26, 1975.
Concepcion, M. The Frustration of Progress: Population Pressures. Proceedings, 4th Annual Seminar for Student Leaders, Manila, Philippines, 1972.
Cooley, W. & Lohnes, P. Multivariate Data Analysis. New York: John Wiley & Sons, Inc., 1971.
Cuyno, R. V.; Ubaldo, Z.; Genio, E.; & Correa, C. "A Non- Formal Education Strategy for Agricultural Development: The Case Study of the Bulacan Farmers Training Center." Paper presented at the symposium on "Rural Development in the Philippines", June 23-24, 1977. UPLB, Philippines.
Dahrendorf, R. Class and Class Conflict in Industrial Society. Stanford, CA: Stanford University Press, 1959.
Dalisay, A. "Institutions in Rural Development." Paper presented at the symposium on "Rural Development Program in the Philippines", June 23-24, 1977, UPLB, Philippines.
Dandekar, V. M. "Comment: Questions of Economic Analysis and the Consequences of Population Growth." In Subsist ence Agriculture and Economic Development. Edited by C. R. Wharton. Chicago: Aldine Pub. Co., 1969. 136
Darlington, R. B.; Weinberg, S. I.: & Walberg, H. I. "Canon ical Variate Analysis and Related Techniques." Review of Educational Research. 43 (1973):433-454.
Erasmus, C . Upper Limits of Peasantry and Agrarian Reform: Bolivia, Venezuela, and Mexico Compared. LTC Reprint No. 38. Madison, W I : Land Tenure Center, 1967.
Griffin, K. Land Concentration and Rural Poverty. London: The MacMillan Press Ltd., 1976.
Falcon, W. F. "The Green Revolution: Generations of Prob lems." Paper presented at the Summer Meeting of the American Agricultural Economists Association. Columbia, Miss.: Aug. 9-12, 1970.
FAO. Training for Agriculture and Rural Development. Rome: Author, 1975.
Felstehausen, H. Conceptual Limits of Development Communi cations Theory. LTC Reprint No. 102. Madison, W I : Land « Tenure Center, 1973.
______& Diaz, H. Communications and Institutional Change in Mexican Agricultural Development. LTC Reprint No 98. Madison, W I : Land Tenure Center, 1974.
Fink, A. & Kosecoff, F. N. An Evaluation Primer. Beverly Hills, CA: Sage Publications, 1978.
Firth, R. "Social Structure and Peasant Economy: The In fluence of Social Structure Upon Peasant Economies." In Subsistence Agriculture and Economic Development. Edited by C. R. Wharton. Chicago: Aldine Pub. Co., 1969.
______. Elements of Social Organization. London: C.A. Watts, 1951.
Flinn, W. L.; Buttel, F. H.: & Havens, A. E. "Attraction and Expulsion as Explanations of Peasant Migration: A Theoretical Analysis and a Columbian Case Study." Paper presented at the Annual Meeting of the Rural Sociolog ical Society, San Francisco, California, Aug. 1975.
Foster, G. M. Traditional Cultures and the Impact of Tech nological Change. New York: Harpers, 1962.
______. "What is a Peasant?" In Peasant Society: A Reader .~Edited by M. Diaz & G. Foster. Boston, MA: Little, Brown and Co., 1967. 137
Frank, A. G. "The Development of Underdevelopment." In Imperialism and Underdevelopment. Edited by R. I. Rhodes. New York: Monthly Review Press, 1970.
George, S. How the Other Half Dies. Monclair, NJ: Allanheld, Osman & Co. Publishers, 1977.
Harding, W. M. "Toward Harnessing Science and Technology for Agricultural Productivity and Community Develop ment." Paper presented at the Third International Semi nar on Communication Strategies in Community Develop ment. Diliman, U.P., Dec. 17, 197 0.
Hays, W. L. Statistics for the Social Sciences. New York: Holt, Rinehart, & Winston, Inc., 1973.
Herdt, R. W. & Barber, R. "Small Rice Farmers and Changing Rice Technology." Journal of Economics and Development. Jan. 1976, VI (1).
Horton, J. "Order and Conflict Theories of Social Problems as Competing Ideologies." The American Journal of Sociology (Undated), pp. 701-713.
Horowitz, I. L. Three Worlds of Development. New York: Oxford University Press, 197 9.
Huizee, G. "The Utilization of Conflict in Community Devel opment and Peasant Organization: A Case Study of Chile." Monograph of the Institute of Social Studies. The Hague, 1972.
Hunter, G . Modernizing Peasant Economies: A Comparative Study in Asia and Africa. London: Oxford University Press, 1969.
______& Bottrall, A. F. (eds.) Serving the Small Farmer: Policy Choices in Indian Agriculture. London: The Chaucer Press, Ltd., 1974.
Igbozurike, M. M. Problem Generating Structures in Nigeria's Rural Development. Uppsala, Sweden: Scandinavian Inst, of African Studies, 1976.
International Labor Office. Poverty and Landlessness in Rural Asia. Geneva: Author, 1977.
Johnston7, B. F. & Kilby, P. Agriculture and Structural Transformation. New York: Oxford Universxty Press, 1975. 138
Joy, J. L. "Comment: Diagnosis, Prediction, and Policy Formulation." In Subsistence Agriculture and Economic Development. Edited by C. R. Wharton. Chicago: Aldine Pub. Co., 1969.
______& Eviritt, E. (eds.) The Rural Problem in North- East Bihar. Sussex: Institute of Development Studies, University of Sussex, 1976.
Kanel, D. Creating Opportunities for Small Farmers: The Role of Land Tenure and Service Institutions. LTC Reprint No. 92. Madison, W I : Land Tenure Center, 1973.
______. Property and Economic Power as Issues in Institu tional Economics. LTC Reprint No. 123. Madison, W I : Land Tenure Center, 1975.
Kennedy, J . J . An Introduction to the Design and Analysis of Experiments in Education and Psychology. Washington, D.C.: University Press of America, 1977.
Kerlinger, F. N. Foundations of Behavioral Research. (2nd ed.) New York: Holt, Rinehart & Winston, Inc., 1973.
Kim, D. H. Small Farmer Economy and Integrated Rural Devel opment . Seoul, Korea: National Agricultural Economics Research Inst., Ministry of Agriculture and Fisheries, 1976.
La Belle, T. J. Nonformal Education and Social Change in Latin America. Los Angeles: UCLA Latin American Center Publications, 1976.
Ladejinsky, W. "Agrarian Reform in Asia, the Green Revolu tion and Its Reform Effects." Paper presented at the 28th International Congress of Orientalists. Canberra, Australia, Jan. 6-12, 1972.
Lenin, V. I. The Development of Capitalism in Russia. Moscow: Foreign Language Publishing House, 1956.
Levin, M. S. Canonical Analysis and Factor Comparison. Beverly Hills, CA: Sage Publications, 1977.
Lewis, 0. Five Families: Mexican Case Studies in the Culture of Poverty. New York: Basic Books, 1959.
Lionberger, H. F. Adoption of New Ideas and Practices. Ames, IW: State University Press, 1961.
Lust, N. "The Relationship of Counselors' Attitudes Toward the Elderly with Selected Counselor Variables." Ph.D. Dissertation. The Ohio State University, 1978. 139
Mann, S. & Dickinson, J. M. "State Agriculture in Two Eras of American Capitalism." Mimeographed copy issued by the Department of Sociology, University of Toronto, Undated.
Marcos, F. E. Notes on the New Society of the Philippines. Manila, Philippines: Author, 1976.
Marx. Theorien uber den Mehrwert, in the translation of G. A. Bonner & E. Burns, Theories of Surplus Value. London: Lawrence & Wishart, 1951.
McAnany, E. G. (ed.) Communications in the Rural Third World: The Role of Information in Development. New York: Praeger, 1980.
Mission to the East. Time, March 2, 1981, 34-36.
Morss, E. R. et al. Strategies for Small Farmer Development. Colorado: Westview Press, Inc., 1976.
Myers, B. "An Empirical Investigation of the Relation of Teacher Perceived Problems by Teacher Personality Variables and Social-Demographic Variables." Ph.D. Dissertation. The Ohio State University, 1977.
Myrdal, G. The Challenge of World Poverty. New York: Pantheon Books, 1970.
Nie, N. et al. Statistical Package for the Social Sciences. New York: McGraw-Hill Book Company, 1975.
Oshima, H. T. "The Strategy of Selective Growth and the Role of Communications." In Communications and Change in Developing Countries. Edited by D. Lerner & W. Schramn. Honululu: East-West Center Press, 1967.
Owen, W. F. "The Significance of Small Farms in Developing Countries." Small Farmer Development Problems. Boulder, CO: CSU, 1974.
Parelman, M. "Author's Reflection." Rural Sociology 44 (1973): pp. 249-251.
Parsons, K. Poverty as an Issue in Development Policy: A Comparison of United States and Underdeveloped Coun tries . LTC Reprint No. 61. Madison, W I : Land Tenure Center, 1969.
______. The Political Economy of Agricultural Devel opment . LTC Reprint No. 116. Madison, WI: Land Tenure Center, 1978. 140
Patrick, G. F.; Brainard, L. J.; & Obermiller, F. W. (eds.) Small Farm Agriculture; Studies in Developing Nations. West Lafayette: Purdue University, 1975.
The Philippines. The Christian Science Monitor. September 19, 1980, B36.
Ponce, E. R. et al. Task Environment Analysis of the First Social Laboratory Barrios in Capiz, Philippines. Capiz, Philippines: CAFS, 1974.
______. Task Environment Analysis of the Three Social Laboratory Barrios in Leyte, Philippines. ViSCA, Baybay, Leyte, Philippines: ViSCA, 1976.
Rhodes, R. I. (ed.) Imperialism and Underdevelopment. New York: Monthly Review Press, 1970.
Rockfeller, J. D. 3rd. "The Challenge of Population and Food." In Subsistence Agriculture and Economic Devel- opment. Edited by C. R. Wharton. Chicago: Aldine Pub. Co., 1969.
Rogers, E. M. "Motivations, Values, and Attitudes of Sub sistence Farmers: Toward A Subculture of Peasantry." In Subsistence Agriculture and Economic Development. Edited by C. R. Wharton. Chicago: Aldine Pub. Co., 1969.
& Shoemaker, F. Communication of Innovations. (2nd ed.) New York: The Free Press, 1971.
______& Burdge, R. J. Social Change in Rural Socie ties. (2nd ed.) New York: Appleton-Century-Crofts, 1972.
Romero, L. K. & Flinn, W. L. "The Differential Impact of Growth on Three Peasant Communities." Paper presented at the Rural Sociological Society Annual Meetings, San Francisco, California, Aug. 30- Sept. 3, 197 8.
Roy, J. "Step-down Procedures in Multivariate Analysis." Annals of Mathematical Statistics. 29 (1958), pp. 1177- 1187.
Rudramoorthy, B. Extension in Planned Social Change. Bombay, India: Allied Publishers Private Limited, 1964.
SAS Institute Inc. SAS User's Guide. 1979 Edition, Cary, NC: Author, 1979.
Schultz, T. W. (ed.) Distortions of Agricultural Incentives. Bloomington, IN: Indiana University Press, 1978. 141
______. Transforming Traditional Agriculture. New Haven: Yale University Press, 1964.
Scott, C. D. Some Problems of Marketing Among Small Peasant Proprietors in Chile. LTC Reprint No. 115. Madison, W I : Land Tenure Center, 1974.
Smock, D. R. & Smock, A. C. Cultural and Political Aspects of Rural Transformation: A Case Study of Eastern Nigeria. New York: Praeger, 1972.
Szal, R. "The Distribution of the Benefits of Growth." G. Standing & R. Szal. Poverty and Basic Needs: Evidence from Guyana and the Philippines. Geneva: International Labor Office, 1979.
Tatsuoka, M. Multivariate Analysis: Techniques for Educa tional and Psychological Research. New York: John Wiley & Sons, 1971.
Thiesenhusen, W. Grassroots Economic Pressures in Chile: An Enigma for Development Planners. LTC Reprint No. 44. Madison, WI: Land Tenure Center, 1968.
______. Green Revolution in Latin America: Income Effects, Policy Decisions. LTC Reprint No. 83. Madison, W I : Land Tenure Center, 1972.
______. Development and Equality: Partners or Rivals. LTC Reprint No. 128. Madison, WI: Land Tenure Center, 1976.
______. Reaching the Rural Poor and the Poorest: A Goal Unmet. LTC Reprint No. 136. Madison, W I : Land Tenure Center, 1978.
Thomas, G. Poverty in Nonmetropolitan South: A Causal Anal ysis . Lexington, KY: D. C. Heath, 1972.
Thorton, D. S. Agriculture in Economic Development. LTC Reprint No. 124. Madison, W I : Land Tenure Center, 1975.
Timn, N. Multivariate Analysis. Brooks: Cole Publishing Company, 1975.
Turner, J . The Structure of Sociological Theory. Homewood, IL: The Dorsey Press, 1978.
United Nations. Economic and Social Survey of Asia and the Pacific, 1975. Bangkok: Author, 1977. 142
Valsan, E. H. Community Development Programs and Rural Local Government; Comparative Case Studies of India and the Philippines. New York; Praeger, 1970.
Villamil, J. Transnational Capitalism and National Develop ment . Sussex, England: The Harvester Press, 1979.
Warwick, P. V. "Canonical Correlation Analysis." In Nie, N. et al. Statistical Package for the Social Sciences. Chicago: McGraw-Hill, 1975.
Watkola, A. "Assessment of the Development, Diffusion and Adoption Package of Agricultural Innovations in Chilalo, Ethiopia." Ph.D. Dissertation. The Ohio State Universi ty, 1975.
Wharton, C. R., Jr. The Green Revolution: Cornucopia or Pandora's Box. Foreign Affairs. April 1969.
______. Subsistence Agriculture and Economic Development. Chicago: Aldine Pub. Co., 1969. A Winkelman, D. "Plan Puebla After Six Years." Proceedings Seminar on Small Farm Agriculture, Purdue University, Sept. 1975.
Wong, J. "An Economic Overview of Agriculture in the People's Republic of China." ADC Teaching Forum. 26, Jan. 1973.
Wood, G. P. "Public Institutions and the Small Farmer." In Small-Farm Agriculture: Studies in Developing Nations. Edited by Patrick et al. West Lafayette, IN: Purdue University, 1975.
Worsley, P. (ed.) Two Blades of Grass. Manchester: Manches ter University Press, 1971.
World Bank. Education. Washington, D. C.: Author, 1974.
Zainuddin, A. P. "Factors Associated with Level of Partici pation of Members of Village Development and Security Committees in Four Penninsular Malaysian States." Ph.D. Dissertation. The Ohio State University, 1977. APPENDIX A
MEANS AND LEAST SQUARE MEANS
143 MEANS
PROVINCE N AGSCORE AVMEAN PARS CORE SUSSCORE
1 - East 100 495.05 47.85 14.59 21.85 2 - West 100 755.99 62.88 15.39 22.02 LAND
1 - Non-owners 121 626.92 56.05 14.79 21.88 2 - Part-owners and 73 623.38 54.32 15.29 22.02 owners CREDIT 0 - No Credit 71 616.98 60.38 15.29 22.40 1 - Bank-Relatives 57 698.99 56.57 15.72 22.89 2 - Usurers 72 575.78 49.47 14.11 20.71 PROVINCE LAND
1 1 49 458.41 45.99 14.57 22.24 1 2 51 530.26 49.64 14.61 21.48 2 1 72 741.60 62.90 14.94 21.63 2 2 28 792.99 62.84 16.53 23.00 PROVINCE CREDIT
1 0 28 487.62 54.58 14.91 22.25 1 1 25 582.51 47.40 15.08 23.47 1 2 47 452.96 44.08 14.14 20.75 2 0 43 701.22 64.16 15.54 22.51 2 1 32 789.98 63.73 16.22 22.43 2 2 25 806.68 59.61 14.05 20.64 LAND CREDIT
1 0 43 623.20 64.14 15.42 23.18 1 1 26 710.51 60.84 14.85 22.23 1 2 52 588.20 46.97 14.24 20.63 2 0 28 607.44 54.60 15.10 21.21 2 1 31 689.32 52.99 16.45 23.44 20 543.49 • 55.97 13.76 20.94 2 2. 144 MEANS
PROVINCE LAND CREDIT N AGSCORE AVMEAN PARSCORE SUSSCORE
1 1 0 14 503.21 60.27 14.90 24.35 1 11 6507.30 46.20 14.33 23.83 1 1 2 29 426.66 39.05 14.46 20.89 1 2 0 14 472.03 48.90 14.92 20.14 1 2 1 19 606.26 47.78 15.31 23.35 1 2 2 18 495.33 52.18 13.62 20.54 2 1 0 29 681.13 66.01 15.67 22.62 2 11 20 771.48 65.23 15.01 21.75 2 1 2 23 791.88 56.96 13.97 20.30 2 2 0 14 742.85 60.31 15.28 22.28 2 2 1 12 820.83 61.24 18.25 23.58 2 2 2 2 976.92 90.13 15.00 24.50 145 LEAST SQUARES MEANS
PROVINCE LAND AGSCORE STD ERR PROB T PROB T HO: LSMEAN(I)=LSMEAN(J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1 2 3 4
1 1 479.06 41.43 0.0001 1 0.3952 0.0001 0.0001 1 2 524.54 33.63 0.0001 2 0.3952 0.0001 0.0001 2 1 748.16 28.39 0.0001 3 0.0001 0.0001 0.1615 2 2 846.87 64.22 0.0001 4 0.0001 0.0001 0.1615 PROVINCE LAND AVMEAN STD ERR PROB T PROB T HO: LSMEAN (I) =LSMEAN (J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1a 2 3 4
1 1 48.50 4.67 0.0001 1 0.8535 0.0129 0.0113 1 2 49.62 3.79 0.0001 2 0.8535 . 0.0090 0.0112 2 I 62.73 3.20 0.0001 3 0.0129 0.0090 0.3243 2 2 70.56 7.24 0.0001 4 0.0113 0.0112 0.3243 PROVINCE LAND PARSCORE STD ERR PROB T PROB T HO: LSMEAN(I)=LSMEAN(J) LSMEAN LSMEAN H0:LSMEAN=0 I/J 1 2 3 4
1 I 14.56 0.41 0.0001 1 0.9176 0.5285 0.0388 1 2 14.62 0.34 0.0001 2 0.9176 . 0.5526 0.0355 2 1 14.88 0.28 0.0001 3 0.5285 0.5526 0.0713 2 2 16.17 0.65 0.0001 4 0.0388 0.0355 0.0713 PROVINCE LAND SUSSCORE STD ERR PROB T PROB T HO: LSMEAN(I)»LSMEAN(J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1 2 3 4
1 1 23.02 0.54 0.0001 1 . 0.0167 0.0260 0.6674 1 2 21.34 0.43 0.0001 2 0.0167 • 0.7145 0.0269 2 1 21.55 0.37 0.0001 3 0.0260 0.7145 0.0395 2 2 23.45 0.83 0.0001 4 0.6674 0.0269 0.0395 LEAST SQUARES MEANS
LAND CREDIT AGSCORE STD ERR PROB T PROB T HO: LSMEAN(I)=LSMEAN(J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1 2 3 4 5 6
1 0 592.17 38.74 0.0001 1 0.4858 0.7380 0.7973 0.0395 0.1388 1 1 639.39 55.41 0.0001 2 0.4858 # 0.6417 0.6550 0.2956 0.3563 1 2 609.27 33.24 0.0001 3 0.7380 0.6417 • 0.9739 0.0598 0.1823 2 0 607.44 44.99 0.0001 4 0.7973 0.6550 0.9739 . 0.0931 0.1975 2 1 713.54 43.89 0.0001 5 0.0395 0.2956 0.0598 0.0931 . 0.8198 2 2 736.12 88.73 0.0001 6 0.1388 0.3563 0.1823 0.1975 0.8198 . LAND CREDIT AVMEAN STD ERR PROB T PROB T HO: LSMEAN(I)=»LSMEAN(J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1 2 3 4 5 6
1 0 63.14 4.37 0.0001 1 0.3315 0.0093 0.2041 0.1930 0.4641 1 1 55.71 6.25 0.0001 2 0.3315 0.2915 0.8904 0.8804 0.1924 1 2 48.00 3.75 0.0001 3 0.0093 0.2915 . 0.2969 0.2960 0.0316 2 0 54.60 5.07 0.0001 4 0.2041 0.8904 0.2969 . 0.9898 0.1420 2 1 54.51 4.95 0.0001 5 0.1930 0.8804 0.2960 0.9898 „ 0.1379 2 2 71.15 10.01 0.0001 6 0.4641 0.1924 0.0316 0.1420 0.1379 . LAND CREDIT PARSCORE STD ERR PROB T PROB T HO: LSMEAN(I)-LSMEAN(J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1 2 3 4 5 6
1 0 15.28 0.39 0.0001 1 . 0.3717 0.0399 0.7633 0.0126 0.3207 1 1 14.67 0.56 0.0001 2 0.3717 . 0.4864 0.5509 0.0037 0.7323 1 2 14.21 0.33 0.0001 3 0.0399 0.4864 . 0.1186 0.0001 0.9226 2 0 15.10 0.45 0.0001 4 0.7633 0.5509 0.1186 . 0.0092 0.4312 2 ’ 1 16.78 0.44 0.0001 5 0.0126 0.0037 0.0001 0.0092 . 0.0147 2 2 14.31 0.89 0.0001 6 0.3207 0.7323 0.9226 0.4312 0.0147 . LAND CREDIT SUSSCORE STD ERR PROB T PROB T HO: LSMEAN(I)=LSMEAN(J) LSMEAN LSMEAN HO:LSMEAN=0 I/J 1 2 3 4 5 6
1 0 23.48 0.50 0.0001 1 . 0.4299 0.0001 0.0037 0.9795 0.4451 1 1 22.79 0.72 0.0001 2 0.4299 . 0.0100 0.0917 0.4631 0.8441 1 2 20.59 0.43 0.0001 3 0.0001 0.0100 . . 0.3994 0.0001 0.1209 2 0 21.21 0.58 0.0001 4 0.0037 0.0917 0.3994 0.0065 0.3142 2 1 2346 0.57 0.0001 5 0.9795 0.4631 0.0001 0.0065 . 0.4644 2 2 22.52 1.15 0.0001 6 0.4451 0.8441 0.1209 0.3142 0.4644 APPENDIX B
INTERVIEW SCHEDULE
148 149
Respondent's Name
FACTORS ASSOCIATED WITH THE AGRICULTURUAL TECHNOLOGY ADOPTION LEVEL, PARTICIPATION LEVEL, AND PERCEPTION LEVEL, OF SAMAHANG NAYON MEMBERS IN LEYTE, PHILIPPINES
INTERVIEW SCHEDULE
Project Code:______Location:______
Household No.______Date:______
Time started:______Time finished:____
Name of Interviewer:______
Signature: ______
* ft "k ie *
(AT DOOR, FROM MEMORY:)
"Good morning/afternoon/eveningi May I speak to Mr./Mrs. ______?"
"Hello, I'm M r . / M s . ______from the Visayas State College of Agriculture, Baybay, Leyte. I would like to ask you some questions, if you like, with regard to your practices, knowledge, and operations in rice farming. In addition, I would like to ask you also some questions regarding your perception and parti cipation in your Samahang Nayon organization. Your answer will be confidential and will be used only to help the college plan better programs for farmers. Your partici pation is voluntary."
"Are you willing to participate?"
(IF YES, PROCEED TO NEXT PAGE) I. AGRICULTURAL KNOWLEDGE ADOPTION LEVEL
(SAY THIS FROM MEMORY:)
"I would like to ask you regarding your rice production practices during the past two growing seasons in which you completed the whole production cycle, i.e., from planting to harvesting and processing.
Answers (Record either in % or Questions fraction)
This past Previous to season last season
1. What fraction or per cent of your field was planted to IRRI or BPI varieties?
2. What fraction or per cent of your field was planted with seeds that were either certi fied/registered or selected from your farm following recommended steps of seed selection?
3. What fraction or per cent of your field was planted with seedlings that were grown using the "dapog" system? __ _
4. What fraction or per cent of your field was planted with straight-row planting? __
5. What fraction or per cent of the missing hills were replanted? __ _
6. What fraction or per cent of the field had basal application of fertilizer?
7. What fraction or per cent of the field was fertilized during vegetation stage? __ 151
This past Previous to season last season
8. What fraction or per cent of the field was fertilized at panicle initiation stage? '______
9. What fraction or per cent of the field was weeded either by chemical or mechanical methods? ______
10. What fraction or per cent of the field was applied with insecticides?
11. What fraction or per cent of the field was irrigated?
12. What fraction or per cent of the harvest was threshed by machine?
13. What fraction or per cent of the harvest was winnowed by machine?
14. What fraction or per cent of the harvest was dried by machine? II. RICE PRODUCTION
(SAY THIS FROM MEMORY:)
"I would like to ask you regarding your gross production, in cavans, from your rice farm during the past two years. By gross production here, I mean your total product without any deduction like share of harvesters, etc.."
January 1978 to December 1978:
Season Area Planted Production (Hectares) (Specify unit of measure) 1. First season. . . . _____
2. Second season . . .
3. Third season. . . .
January 1979 to December 1979:
Season Area Planted Production (Hectares) (Specify unit of measure)
1. First season. . . .
2. Second season . . .
3. Third season. . . . III. PARTICIPATION LEVEL
(SAY THIS FROM MEMORY:)
"I would like to ask you regarding your partici pation in the following activities of your Samahang Nayon for the last twelve months. For each activity, please rate yourself with any of the following responses which best describe your participation.
1 - Very poor (25% attendance or l.ess) 2 - Poor (21% to 40% attendance) 3 - Fair (41% to 60% attendance) 4 - Good (61% to 80% attendance) 5 - Very good (81% to 100% attendance)
Activities: Rating (CHECK ONE) 1 2 3 4
1. Attendance in meetings. . . ______
2. Attendance in extension classes to learning new agricultural technology . ______
3. Attendance in cooperative work like repair of dikes, road, e t c ......
4. Attendance in social acti vities like dance, athletic games, etc . . . IV. PERCEPTION LEVEL
(SAY THIS FROM MEMORY:)
"I would like to ask you regarding some of your perceptions (outlook) regarding the future of your Samahang Nayon organization. Each of the following state ments have five possible responses. Please select one of the responses which best describes your opinion or feeling. The possible responses are:
1 - Strongly disagree
2 - Disagree
3 - Uncertain
4 - Agree
5 - Strongly agree
Statements Responses (CIRCLE ONE ONLY)
1. My Samahang Nayon will continue to exist and operate without the help of its present MDO technician . . 1
2. Should there be no BAEx technician I am willing to help pay for a private technician who will give me assistance in farming...... 1
3. Within the next five years, I pro ject that in cooperation with other Samahang Nayon organizations my organization will be engaged in wholesale buying of farm inputs like fertilizers and chemicals for its members ...... 1
4. Within the next five years, I pro ject that in cooperation with other Samahang Nayon organizations my organization will be engaged in cooperative marketing or rice pro duced by its members...... 1 Within the next five years, my Samahang Nayon organization will guarantee rural bank loans for its members......
Within the next five years, my Samahang Nayon organization will grow in membership ......
Within the next five years, my Samahang Nayon organization will have weaker influence in the economic and political activi ties of my village ...... V. GROSS INCOME
Please indicate your major and minor sources of income.
(AFTER ASKING THE RESPONDENT HIS/HER MAJOR AND MINOR SOURCES OF INCOME, PLEASE LIST THEM BY CATEGORIES (MAJOR OR MINOR SOURCES) IN THE APPROPRIATE COLUMNS ON NEXT PAGE.
FOR EVERY INCOME SOURCE, PLEASE ASK THE RESPONDENT IF IT IS RECEIVED WEEKLY OR MONTHLY AND RECORD THE AMOUNT OR QUANTITY IN THE CORRESPONDING COLUMN. RECORD FARM PRODUCE IN CATEGORIES, SUCH AS RICE, ROOTCROPS, BANANA, VEGETABLES, HOGS, ETC. PLEASE INQUIRE IF PRODUCE IS HARVESTED WEEKLY, MONTHLY, ETC, RECORD THE QUANTITY HARVESTED IN TERMS OF SACKS (CAVANS), CANS, KILOS, PIECES, ETC. VALUE OF PRODUCE IS BASED ON CURRENT MARKET PRICE.
WITH CASH INCOME OR PRODUCE THAT OCCURS LONGER THAN A MONTH, DIVIDE THE AMOUNT OR QUANTITY BY THE NUMBER OF MONTHS IN WHICH THE INCOME OCCURS TO GET THE MONTHLY INCOME. TO COMPUTE FOR YEARLY INCOME IN BOTH MAJOR AND MINOR SOURCES, RESPECTIVELY, MULTIPLY EACH BY 12) 157 a) Major Sources:
Cash Income or Amount or Quantity Received or Harvested: Produce Weekly Weekly Monthly Monthly amt./qnty. Value amt./qnty. Value —y — y
Total Major Sources: Monthly: f
Yearly: f b) Minor Sources;
Cash Income or Amount or Quantity Received or Harvested: Produce Weekly Weekly Monthly Monthly amt./qnty. Value amt./qnty. Value t ?
Total Minor Sources: Monthly: f
Yearly: f 158
VI. OTHER CRITERION VARIABLES
A. Demographic Characteristics
1. Age (of last birthday) ______
2. Sex: Male _____ Female
3. Civil status: _____ Married _____ Widow(er) _____ Single
4. Educational attainment (highest grade level) ______
(IF RESPONDENT HAS NOT GONE TO SCHOOL, PLEASE ASK IF HE/SHE CAN READ AND WRITE)
4a. Can you read and write? _____ Yes______No
5. At present, how many people, besides you, live in your house? What are their names, relationship to you, ages, sex, and occupation?
Relation Name to you Age Sex Occupation
1) ______;______; ______
2) ___
3 ) ...... '______
4) ______
5 ) ______
6) ___ ; _
7 ) ______
8) ; ______; ______
9 ) ______
1 0 ) ______ID ______6. Total: (Number of persons) 159
7. At present, do you own the land from which you derived most of your income?
Yes (GO TO Q. 9)
_____ Partly No
8. If no or partly, what type of arrangement are you able to use it?
(PROBE ANSWER AND SPECIFY THE RENT OR SHARE OF THE LANDLORD EITHER IN f OR IN COMMODITY)
_____ Leaseholder/rental Rent per season______
Share system/ — * Share of land - tenancy lord per season
9. How many years have you been farming? ______
10. How many years are you a Samahang Nayon member?
11. Do you presently avail of credit to finance your farm operation?
(CHECK ALL THOSE THAT APPLY)
_____ No
Yes, from the bank
_____ Yes, from a money lender
Yes, from a relative
Others, specify: ______160
B. Technical Factors
1. How would you rate the technical competence of the Bureau of Agricultural Extension technician in relation to his/her ability to help you tackle/solve technical problems of your farm?
Very p o o r ......
Poor......
Neither good nor poor . . . '
Good......
Very g o o d ......
2. How would you rate the technical ability of your MDO technician to help your Samahang Nayon?
(CHECK ONE)
Very poor
Poor
Neither good nor poor
Good
Very good
3. How would you rate the communication flow between the MDO technican and the members of your Samahang Nayon? (CHECK ONE)
Very poor
Poor
Neither good nor poor
Good
Very good
4. How often do you meet with the technicians during the past three months? MDO BAEx Number of times...... 161
5. Have you been involved in the idea evolution and design of your Samahang Nayon? (CHECK ONE)
No
Rarely
Sometimes
Most of the time
All the time
6. At present, are you involved in making decisions regarding the operation of your Samahang Nayon?
(CHECK ONE)
No
_Rarely
Sometimes
Most of the time
All the time
C. Political Factors
How would you judge the support and cooperation of the following people in terms of achieving the objectives of your Samahang Nayon? (MENTION PERSONS LISTED BELOW, PROBE ANSWER/ AND CIRCLE APPROPRIATE CATEGORY)
Persons: Circle one: Very Neither Very weak Weak weak nor Strong strong strong
Local barrio officials . .
Town mayor and councilmen. . 162
V. PHYSICAL FACTORS
Project Code: ______Date:
Household No. MDO Technician:
Location: (Address)
(FOR EACH RESPONDENT SELECTED, KINDLY SECURE THE NECESSARY INFORMATION FROM THE SOURCES INDICATED)
1. Physical Constraints: From actual visitation. (PLEASE GET THE EVALUATION OF THE PROPER PERSONNEL OF AGENCY CONCERNED THAT ARE STATIONED IN THE TOWN/ DISTRICT ON THE FOLLOWING PHYSICAL ASPECTS OF THE RESPONDENT'S FARM USING THE CODE BELOW:)
1 - Very poor 2 - Poor 3 - Good 4 - Very good
_____ Soil fertility (Ask Bureau of Soils or BAEx technician)
_____ Irrigation sufficiency (Ask National Irrigation Administration or BAEx technician)
_____ Farm to market condition (Ask Bureau of Highways or personal assessment of inter viewer)
_____ Flood control adequacy (Ask Bureau of Highways or other technicians stationed in the area.)
2. Market Access (Ask respondent or MDO technician)
Average distance in kilometers to the nearest all-weather road
Average distance in kilometers to the nearest town market