 

I?!
.. :2
.1

 

‘ r.
. . .mﬁhmﬂ

 

. lid
‘
. .3
Sunk.
«1..
z

w

 

 

J}. .1. .utzlh.

 

 

«1‘33 3
u. , .’

3 . LIBRARY
k w Michigan State
.2 3 J4, University

This is to certify that the
thesis entitled

FARMERS’ ATI'ITUDES AND ADOPTION OF IMPROVED MAIZE
VARIETIES AND CHEMICAL FERTILIZERS IN THE MANICA
DISTRICT, MOZAMBIQUE

presented by

Eunice Paula Armando Cavane

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Community. Agriculture,
Recreation and Resource Studies

\ Major Pr7ess s Signaturf
l2, \‘tI 0?

Date

 

 

MSU is an aﬁinnative-ection, equal-opportunity employer

_ ﬁ.—l-.---o-nan-o-r-

—.—.-.-.-.-.-.-.-—.- _

—»—.- ---o-n--I-o-o-o-.—.- _

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/ClRC/DateDue.indd-p.1

FARMERS’ ATTITUDES AND ADOPTION OF IMPROVED MAIZE VARIETIES
AND CHEMICAL F ERTILIZERS IN THE MANICA DISTRICT, MOZAMBIQUE

By

Eunice Paula Armando Cavane

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Community, Agriculture, Recreation and Resource Studies

2007

ABSTRACT

FARMERS’ ATTITUDES AND ADOPTION OF IMPROVED MAIZE VARIETIES
AND CHEMICAL FERTILIZERS IN THE MANICA DISTRICT, MOZAMBIQUE

By
Eunice Paula Armando Cavane
Improved maize varieties and chemical fertilizers nitrogen, phosphorus, and
potassium (12-24-12 NPK) and urea have been disseminated in Mozambique to raise the
productivity of maize, a major staple crop. This study determined the inﬂuence of
farmers’ characteristics and farmers’ attitudes toward improved maize and chemical
fertilizers. Rural households producing maize in the highlands of Machipanda and
lowlands of Vanduzi represented the population of interest in this investigation. One
hundred and twenty (n = 120) households from Machipanda and one hundred and
seventy-three (n = 173) households from Vanduzi were randomly selected for the study.
Data were collected through personal interviews. The results showed that the level of
education was low among the respondents, particularly in the lowlands of Vanduzi. Most
farmers were knowledgeable of the advantages and disadvantages of using improved
maize varieties in Machipanda and Vanduzi, but few respondents were knowledgeable
about application methods of NPK and urea fertilizers on maize, particularly in the
lowlands of Vanduzi. In both study areas, respondents tended to hold a generally positive
attitude toward the marketability of improved maize variety SC513, production
characteristics of SC513, and use of NPK and urea for maize production. However, the
strength of farmers’ attitudes toward improved maize variety SC513 and chemical
fertilizers NPK and urea varied according to the location (Machipanda and Vanduzi) and

within location according to gender and sources of information (neighbors, extension,

and market). Male respondents tended to hold stronger positive attitudes toward
production characteristics of improved maize variety SC513 and use of NPK and urea for
maize production. The sources of information did not affect attitudes of farmers in
Machipanda. Only in Vanduzi, farmers who learned about chemical fertilizers from
extension services tended to hold stronger positive attitudes toward chemical fertilizers
than farmers who learned about these technologies from neighbors. Adoption of
improved maize variety SCS 13 was higher than the adoption of chemical fertilizers.
How-to knowledge and agro-ecological region were the common factors in adoption of
the improved maize variety and chemical fertilizers. In particular, attitude toward
production characteristics and attitude toward marketability of improved maize variety
were important factors in adoption of maize variety. Extension services were important
factors of adoption of chemical fertilizers. After farmers had adopted the technologies
for one or more years, discontinuance occurred mainly because of lack of money to
purchase seed and fertilizers, susceptibility of grain to attack by storage weevils, and non-
complete closure of husk cover, which exposed the cob to rain and attack by ﬁeld insects
and diseases. The National Directorate of Rural Extension, the private sector, and
research institutions should expand their representatives to the local level and coordinate
knowledge-led agricultural development with a focus on ﬁnding varieties that have wide
adaptation, growing well in the highlands and lowlands; improving how-to knowledge on
fertilizer application; prioritizing production and marketability characteristics of
improved maize varieties; providing fertilizers and seed on a credit basis; promoting
women extension workers and input suppliers; and strengthening farmers’ positive

attitudes toward improved maize varieties and chemical fertilizers.

DEDICATION

To my parents, Armando Cavane Mutacate and Paulina José Chichava, for their love,
support, and encouragement.

To my husband, Ilidio Z. Goenha, and my children, Mércio L. Goenha and Nicole
Goenha, for your patience and loving support.

To my younger brothers, Henrique, Marcelo, Se’rgio, José, Armando, Silvino, and
Francisco, and my sisters, Felismina and Sabina, for your encouragement.

iv

ACKNOWLEDGMENTS

I am especially grateful to my major professor, Dr. Murari Suvedi, for his
excellent supervision and guidance throughout the years of my studies. My deep
gratitude goes to the advisory committee members, Dr. Eddie Moore, Dr. Cynthia
Donovan, and Dr. Russell Freed. I am especially grateful for your concern, patience, and
support of my professional development.

I am deeply grateful to the Ford Foundation, Larenstein School, and the
Department of Community, Agriculture, Recreation and Resource Studies for providing
the ﬁnancial support for the completion of my studies at Michigan State University. I
would like to express my deep appreciation to Michigan State faculty members for
providing a rich learning environment.

I must express my special appreciation to farmers in Machipanda and Vanduzi
administrative posts of the Manica district for their warm hospitality, unconditional
collaboration, and profound commitment to this study. I also would like to thank Tom
Walker for his contributions on the survey instrument and Ms. Maria de Luz Quinhetos
for her outstanding help in data collection.

I am grateful to the Faculty of Agronomy and Forestry Engineering of Eduardo
Mondlane University (FAEF), particularly the Unit of Communication and Rural
Sociology and Mr. Afonso Nhoela for providing assistance during ﬁeldwork.

Special thanks to André Boon, Catherine Vijfhuizen, Bart Pijnenburg, Luisa
Santos, Firmino Mucavele, Roland Brown, Antoinette Vogt, and Célia Dinis, for

believing in me.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................. viii
LIST OF FIGURES ................................................................................ x
CHAPTER I
INTRODUCTION ................................................................................. 1
Statement of the Problem ......................................................................... 3
Purpose of the Study .............................................................................. 5
Importance of the Study ........................................................................... 6
Deﬁnition of Terms ................................................................................ 7
Organization of the Study ........................................................................ 9
CHAPTER 11
REVIEW OF RELATED LITERATURE ..................................................... 10
Section 1. Agricultural Sector in Mozambique ............................................... 10
Section 2. Beneﬁts and Disadvantages of Using Improved Maize and Chemical
Fertilizer ........................................................................................... 22
Section 3. Deﬁnitions and Theoretical Views about Attitudes ............................. 25
Section 4. Empirical Studies on Farmers’ Attitudes Toward Agricultural
Technology ....................................................................................... 31
Section 5. Theoretical Measurement Models for Development of Attitudinal
Scales ............................................................................................................................. 34
Section 6. Adoption and Diffusion Theory .................................................... 37
Section 7. Theoretical Framework of Adoption of Hybrid Maize SC513 and Chemical
Fertilizers NPK and Urea ........................................................................ 45
Section 8. Study Hypotheses .................................................................... 53
Section 9. Logistic Model ....................................................................... 54
CHAPTER III
METHODOLOGY ............................................................................... 57
Section 1. Study Area ........................................................................... 57
Section 2. Research Design ..................................................................... 64
Section 3. Data Examination ................................................................... 71
Section 4. Data Analysis ......................................................................... 76
Section 5. Study Limitations .................................................................... 79
CHAPTER IV
RESULTS .......................................................................................... 81
Research Objective 1 ............................................................................ 81
Research Objective 2 ............................................................................ 85
Research Objective 3 ............................................................................ 93
Research Objective 4 ............................................................................ 99
Research Objective 5 ........................................................................... 105

vi

CHAPTER V

DISCUSSION .................................................................................... 1 13
CHAPTER VI

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ........................ 135
APPENDICES .................................................................................. 151
Appendix A: Map of Mozambique ........................................................... 152
Appendix B: Maize Technology Package, Central and Northern Mozambique ....... 154
Appendix C: Letter of Consent ............................................................... 156
Appendix D: Survey Instrument ............................................................... 158
Appendix E: Exploratory and Conﬁrmatory Factor Analysis ............................ 178
Appendix F: Collinearity Diagnostics and Logistic Regression .......................... 183
REFERENCES .................................................................................. 191

vii

LIST OF TABLES

Table 1. Seed and Fertilizer Levels to Produce 4 tons of Maize Grain per Hectare ................ 15
Table 2. Characteristics of Improved Maize Hybrid Variety SC513 ....................................... 16
Table 3. Average Prices of Inputs and Maize Grain ................................................................ 21
Table 4. Budget for Hybrid Maize SC 513 .............................................................................. 21
Table 5. Budget for Local Maize ............................................................................................. 21
Table 6. Partial Budgeting: Change from a Local to Hybrid Maize ........................................ 22
Table 7. Key Agro-climatic and Farming Characteristics of Machipanda and Vanduzi ......... 63

Table 8. Multistaging Sampling of Households in Machipanda and Vanduzi Administrative

Posts ................................................................................................................................. 65
Table 9. Structure of the Survey Instrument ............................................................................ 66
Table 10. Factor Loading and Coefficients of Reliability on Attitudes Toward SC513 ......... 73
Table 11. Factor Loadings and Coefﬁcient of Reliability on Attitudes Toward NPK ............ 74
Table 12. Factor Loadings and Coefﬁcient of Reliability on Attitudes Toward urea ............. 75
Table 13. Fit Statistics for Conﬁrmatory Analytic Models ..................................................... 75
Table 14. Description of Variables Entered in the Multiple Logistic Regression ................... 79
Table 15. Characteristics of Respondents in the Machipanda and Vanduzi Administrative

Posts in the Manica District ............................................................................................. 82
Table 16. Respondents’ Attitudes Toward Improved Maize Variety SC513 .......................... 91
Table 1?. Respondents’ Attitudes Toward NPK ...................................................................... 92
Table 18. Respondents’ Attitudes Toward Urea ...................................................................... 92
Table 19. ANOVA Table for One-Way Analysis of Mean Attitude by Age Categories ........ 95
Table 20. ANOVA Table for One-Way Analysis of Mean Attitude by Sources of

Information ...................................................................................................................... 96
Table 21. Mean Attitudes by Gender of Respondents ............................................................. 97

viii

Table 22. Mean Attitudes by Study Location .......................................................................... 98

Table 23. Percentage of Farmers Who Ever Used and are Currently Using the
Technology .................................................................................................................... 102

Table 24. Reasons for Discontinuance of Use of Hybrid Maize SC513 Among Farmers in
Machipanda and Vanduzi .............................................................................................. 103

Table 25. Reasons for Discontinuance of NPK Use Among Farmers in Machipanda and
Vanduzi .......................................................................................................................... 104

Table 26. Reasons for Discontinuance of Urea Use Among Farmers in Machipanda and
Vanduzi .......................................................................................................................... 104

Table 27. Maximum Likelihood Estimates of the Logistic Model for Factors Affecting
Adoption of SC513 Among Respondents in Machipanda and Vanduzi ........................ 107

Table 28. Maximum Likelihood Estimates of the Logistic Model for Factors Affecting
Adoption of NPK Among Respondents in Machipanda and Vanduzi .......................... 1 10

Table 29. Maximum Likelihood Estimates of the Logistic Model for Factors Affecting
Adoption of urea Among Respondents in Machipanda and Vanduzi ............................ 111

ix

LIST OF FIGURES

Figure 1. Distribution Channels for Improved Maize Seeds in Mozambique ......................... 18
Figure 2. Distribution Channels for Chemical Fertilizers in Mozambique ............................. 19
Figure 3. The Latent Process Viewpoint of Attitudes. Adapted from Oskarnp and Schultz
(2005) and Eagly and Chaiken (1993) ............................................................................. 27
Figure 4. Technology Adoption Framework (N eupane et. al., 2002) ...................................... 42
Figure 5. The Diffusion Process Represented by the S-shaped Curve .................................... 44
Figure 6. Conceptual Framework for Adoption of Hybrid Maize SC513, NPK and Urea ...... 47
Figure 7. Map of the Manica District Showing the Five Administrative Posts, including the
Two Research Sites, MACHIPAN DA and VANDUZI Administrative Posts ................ 58
Figure 8. The Unimodal Pattern of Rainfall in the Manica District ........................................ 59

Figure 9. Pattern of Adoption of Hybrid Maize SC513 in Machipanda and Vanduzi from
1990 to 2005 .................................................................................................................. 100

Figure 10. Pattern of Adoption of Fertilizer NPK in Machipanda and Vanduzi from 1977 to
2005 ............................................................................................................................... 100

Figure 11. Pattern of Adoption of Fertilizer Urea in Machipanda and Vanduzi from 1972 to
2005 ................................................................................................................................ 101

CHAPTER I

INTRODUCTION

The Republic of Mozambique lies on the eastern coast of southern Africa and has
an area of approximately 800,000 square kilometers (Appendix A). The size of the
population is approximately 19 million The bordering countries are Tanzania to the
north; Malawi, Zambia, and Zimbabwe to the west; and South Aﬁ'ica and Swaziland to
the south. Mozambique is divided into 10 provinces: Maputo, Gaza, and Inhambane
provinces in the south; Sofala, Manica, Tete, and Zambezia in the center; and Nampula,
Cabo-Delgado, and Niassa provinces in the north.

Nearly 80 percent of the population lives in rural areas and is engaged in
agricultural production. Approximately 99 percent of the farming population is
smallholder farmers whose production systems are heavily dependent upon rainfall.
Smallholder farmers rarely use purchased inputs, and their production is used to sustain
their households. Subsistence farming does not provide farmers with the surplus
necessary to develop market-oriented agriculture, which could help farmers improve their
livelihoods. Consequently, the levels of productivity are low, markets are poorly
developed, the overall agricultural sector is undeveloped, and absolute poverty is still
affecting the majority of the population. About 65 percent of Mozambicans live on less
than US $1 per day (Programa Nacional para Desenvolvimento Agrario [PROAGRI],
(Agriculture Development National Programme), 2003), making it difﬁcult to meet basic

needs such food, shelter, clothing, education, and sanitation.

Reducing rural poverty became one of the most pressing economic and social
issues in the Republic of Mozambique. Mozambique’s Action Plan for the Reduction of
Absolute Poverty (PARPA) and the Agriculture Development National Programme
(PROAGRI) are among the main policy instruments developed to assist in the ﬁght
against poverty. The PARPA sets the general strategy for poverty reduction and provides
the main contributions by each public sector. The PROAGRI sets the agricultural
sector’s strategy for ﬁghting poverty. PROAGRI’s main goal is to assist farmers to
transition from subsistence farming to market-oriented agriculture (PROAGRI, 2003).

The PROAGRI and the Action Plan for the Reduction of Absolute Poverty
(PARPA) see adoption of improved agricultural technologies by farmers to increase
agricultural production as key in the ﬁght against poverty. According to PARPA,
farmers need increased access to extension and research, and smallholder farmers should
adopt improved production technologies to improve productivity. The PARPA predicted
that by the year 2003, about 460,000 out of 3 million farmers would have adopted
improved agricultural technologies and productivity would have increased by more than
200 percent (PARPA, 2001). The PROAGRI stresses the dissemination of improved
technology and crop management practices for principal food crops, including maize
(Zea mays L.).

In Mozambique, maize is the most cultivated food crop and is a traditional food
staple. It is cultivated by approximately 80 percent of rural households and is planted on
about 40 percent of the cultivated land (Instituto Nacional de Estatistica [INE], (National
Institute of Statistics), 2001). Maize yields are low, averaging between 0.4 and 1.3 tons

per hectare (Jeje et al., 1998). Non-adoption of improved varieties and chemical

fertilizers is one reason farmers are obtaining low maize yields. One major consequence
of low maize yield is that the country’s objective of food security remains
unaccomplished and the country is highly dependent on food imports, in spite of high
agricultural potential (Bias and Donovan, 2003).

This study explored how the decision to adopt hybrid maize varieties and use
chemical fertilizers nitrogen, phosphorus, and potassium (12-24-12 NPK) and urea to
increase average maize yield is inﬂuenced by farmers’ attitudes toward improved maize
and chemical fertilizers and variables such as agro-ecological zone, age, family size,
education, how-to knowledge, and sources of information on improved maize

technology.

Statement of the Problem

One of the challenges in creating agricultural development policies in
Mozambique is achieving widespread adoption of agricultural technologies aimed at
increasing the productivity of maize. Despite availability of improved maize varieties
and chemical fertilizers to improve maize production (J eje et al., 1998), these
technologies are still not widely adopted in Mozambique (INE, 2001; Uaiene, 2004). The
majority of farmers are still relying on saved seed and do not use chemical fertilizer for
maize production. The cultivation of improved maize varieties and application of
chemical fertilizers to increase productivity will continue to be a policy focus for years to
come because the Mozambican government is determined to promote a "Green
Revolution" as outlined by the president’s discourse on agricultural development
(Agéncia Mocambicana de Informacao [AIM] (Mozambican News Agency), Reports No.

337, March 20, 2007):

...the agriculture ministry has the vital task of leading a Mozambican
"Green Revolution" to ﬁght against hunger and to create more jobs.
Among the factors involved in a Green Revolution, ..., are the
production of improved seeds, the supply of fertilizers, the
improvement of irrigation systems, and the transformation of
subsistence farmers into commercial farmers. (Mozambique News
Agency, AIM Reports No. 336, March 5, 2007).

This discourse not only expresses political support for adoption of improved
agricultural technology but also indicates that a greater emphasis on dissemination of
information on improved agricultural technology is expected to be seen. Knowledge of
farmers’ adoption of improved technologies is vital to create agricultural development
policies and implement future technology dissemination programs. However, research to
further advance our understanding of farmers’ apparent unwillingness to adopt improved
maize varieties and chemical fertilizers is still limited.

But there are opportunities to learn more about improved maize and chemical
fertilizers in Mozambique. Since 1995, the National Directorate of Rural Extension and
Sasakawa Global 2000 program (or DNER/SG2000) has been disseminating improved
maize varieties (open-pollinated varieties and hybrids) and chemical fertilizers (N PK and
urea) for improving maize production, resulting in an increase in potential maize yield
from 0.5 ton/ha to 6 ton/ha (DNER, 2002). In the Manica district, the program was
implemented in Machipanda and Vanduzi from 1996 to 2001 (personal communication,
with head of Manica District Public Extension, 2005). From 1996 to 2001, the
DNER/SG2000 program distributed a technological package on the new Hybrid and
fertilizer technology, which included the improved maize variety Hybrid SC513 and NPK

12-24-12 and urea (DNER, 2002). According to Gemo, Eicher, and Teclemariarn,

(2005), the DNER/SG2000 program had encouraged farmers to adopt improved seed
varieties and chemical fertilizers. Therefore, the opportunity is there for researchers to
conduct studies that will shed light on how farmers came to adopt improved maize
varieties and fertilizers.

The inﬂuence of variables such as household size, hired labor, years of
experience, and years of education on adoption of improved maize varieties and chemical
fertilizers has been extensively studied in other African countries (F eder, Just, and
Zilberman, 1985; Kaliba, Verkuijl, and Mwangi, 2000; Abebaw and Belay; 2001;
Tesfaye and Alemu, 2001; Tesfaye, Bedassa, and Shiferaw, 2001). Though not as
extensively observed, attitudes are also considered explanatory of adoption behavior
(Rogers, 1995; Chilonda and Van Huylenbroeck, 2001). Negative attitudes toward
improved maize varieties and chemical fertilizers may be one of the reasons farmers are
not using them for maize production in Mozambique.

This study investigates the adoption of the improved maize variety SC513 and
NPK and urea chemical fertilizers by farmers in the Machipanda and Vanduzi
administrative posts of the Manica district. The study focuses on two aspects: the
farmers’ attitudes about and the factors associated with the adoption of the improved

maize variety SC513 and chemical fertilizers.

Purpose of the Study

The overall purpose of this study was to determine the inﬂuence of farmers’
characteristics and farmers’ attitudes toward improved maize and chemical fertilizers on

adoption of improved maize SC513 and NPK and urea fertilizers by farmers in the

highlands of Machipanda and lowlands of Vanduzi administrative posts of the Manica

district. The speciﬁc objectives were:

1.

To describe the characteristics of respondents from Machipanda and Vanduzi
administrative posts.

To assess the attitudes of farmers in the Machipanda and Vanduzi administrative
posts toward the improved maize variety SC 5 1 3 and chemical fertilizers NPK and
urea.

To compare mean attitude scores between Machipanda and Vanduzi
administrative posts, and among the age categories, sources of information, and
gender of respondents within Machipanda and Vanduzi administrative posts.

To describe the pattern of adoption of the improved maize variety SC513 and
chemical fertilizers NPK and urea in Machipanda and Vanduzi administrative
posts.

To determine the factors associated with adoption of the improved maize variety
SC513 and chemical fertilizers NPK and urea in Machipanda and Vanduzi

administrative posts.

Importance of the Study

The study provides valuable information on pattern of adoption as well as factors

associated with the adoption of the improved maize variety SC513 and chemical

fertilizers among small farmers in the Machipanda and Vanduzi administrative posts.

The study assists in the creation of agricultural deve10pment policy and the

implementation of future technology dissemination programs. In particular, the results of

this study are relevant for the National Directorate of Rural Extension (DNER) and the

national and international research institutions that collaborate with the DNER for the
dissemination of improved maize varieties and chemical fertilizers. The study formulates
practical implications for dissemination of improved maize varieties and chemical

fertilizers in the study area.

Deﬁnition of T erms

Attitude: A predisposition to respond in a favorable or unfavorable manner with respect
to a given attitude object (Oskamp and Schultz, 2005). “Attitude object” includes

situations, institutions, concepts, or persons (Aiken, 2002).

Improved maize varieties: Hybrids and/or open-pollinated varieties (OPV) (Byerlee,
1994) that have been improved for selected traits such as drought tolerance, disease
resistance, pest resistance, grain texture, short maturity rate, increased yield per unit of

land, and quality protein (Denic, 2005).

Chemical fertilizer (N PK 12-24-12): Dry fertilizers consisting of the three primary
macronutrients — nitrogen (N), phosphorus, (P) and potassium (K) - to increase soil
fertility (Hoeft, Nafziger, Johnson, and Aldrich, 2000). NPK is usually placed in the soil
at planting time.

Chemical fertilizer urea: Urea is the most widely used dry nitrogen (N) fertilizer and is
usually side-dressed when the crop is knee-high (Hoeft, Nafziger, Johnson, and Aldrich,
2000)

Adoption: The cultivation of improved maize varieties and application of chemical

fertilizers (NPK and urea) for maize production in a given period of time.

Adoption rate: The relative speed with which an innovation is adopted by members of a
social system. Adoption rate is measured as the percentage of adopters in a given period

of time, for example, a year (Rogers, 1995).

Rain-fed farming: Growing crops or animals under conditions of natural rainfall (Beets,

1990)

Province: A province is an administrative subdivision of the country. Each province is

subdivided into several districts.

District: A district is an administrative subdivision of a province. Each district is

subdivided into several administrative posts.

Administrative post: An administrative post is a subdivision of the district. Each
administrative post is subdivided into several localities, which are further subdivided into

several villages.

Sasakawa Global 2000: SG 2000 is the joint program of the Sasakawa African
Association, from the Nippon Foundation, and Global 2000, from the Carter Center. The
SG2000 began the diffusion of improved agricultural technology to small-scale farmers
in 1986 in Ghana and Sudan, then spread to more than 10 Sub—saharan countries

(Sassakawa Africa Association, 2007).

Soil nutrient mining; Deterioration in soil physical, chemical, and biological properties.

It occurs through a combination of lowering of soil organic matter and loss of nutrients.

HA: Trabalho the Inquérito Agricola (Agricultural Survey).

Organization of the Study

This study is organized in six chapters. Chapter One presents the introduction,
problem statement, objectives of the study, importance of the study, and deﬁnition of
terms. Chapter Two begins with a description of the agricultural sector and
dissemination of improved maize production technology in Mozambique. The
description is followed by the presentation and discussion of literature on the beneﬁts and
risks of improved maize varieties and chemical fertilizers, attitude formation, the
relationship between attitude and overt behavior, and the innovation-decision process.
Chapter Three presents the methodology used for collecting, examining, and analyzing
data. Chapter Four presents the results of the study. Chapter Five presents the
interpretation and discussion of the results. Chapter Six presents the summary,

conclusions, implications of the study, and recommendations for further research.

CHAPTER 11

REVIEW OF RELATED LITERATURE

This chapter is organized in nine sections. Section 1 describes the agricultural
sector and dissemination of improved maize varieties and chemical fertilizers in
Mozambique. Section 2 presents the beneﬁts and disadvantages of using improved maize
varieties and chemical fertilizers. Section 3 reviews the deﬁnitions and theoretical views
of attitudes. Section 4 presents empirical studies on farmers’ attitudes toward agricultural
technology. Section 5 identiﬁes attitudinal dimensions (factors) based on literature
review on beneﬁts and disadvantages of improved maize varieties and chemical
fertilizers, and literature review on attitude formation. Section 6 describes various
theoretical procedures for studying adoption of agricultural technologies. Section 7
describes the theoretical framework used in this study for studying the adoption of the
improved maize variety SC513 and chemical fertilizers. Section 8 presents the study
hypotheses on the relationship between factors and adoption of SC513 and chemical

fertilizers. Section 9 presents the logistic regression model.

Section 1. Agricultural Sector in Mozambique

Agricultural Enterprises

In Mozambique, the agricultural sector is characterized by three categories of
agricultural enterprises (or explorac‘o'es) - small, medium, and large enterprises. This
classiﬁcation is based on the area cultivated and the number of animals owned. Thus,

farms with cultivated areas of less than 10 hectares and with fewer than 10 head of cattle,

10

50 goats/sheep/pigs, and/or 5,000 poultry are classiﬁed as small-scale farms. Medium-
scale farms are those with 10 to 50 hectares of cultivated area, between 10 and 100 head
of cattle, between 50 and 500 goats/sheep/pigs, and/or between 5,000 and

20,000 poultry. Large-scale farms are any that have one or more than one component
larger than the medium-scale limit. For irrigated land, horticultural crops, and
plantations, small-scale is less than 5 hectares, medium is 5 to 10 hectares; and large-
scale is over 10 hectares (INE, 2001; Bias and Donovan, 2003).

Agra-ecological Regions

Ten distinct agro-ecological regions are found in Mozambique. The regions R1,
R2, and R3 are located in southern Mozambique. These regions are subject to prolonged
periods of drought. The annual rainfall is between 500 and 800 mm. High levels of
precipitation are observed in regions R4 to R10, located in the northern part of the
country. The annual rainfall in regions 4-10 is between 1,000 and 1,500 mm.

In general, Mozambique’s soils are considered of average quality, with better
soils found in the plateau and highland of northern and western Mozambique (Jeje et al.,
1998). However, it is important to note that negative nutrient balance of the soils and
consequent gradual loss of soil productivity have been observed in cultivated areas in
Mozambique (Instituto Nacional de Investigacao Agraria [INIA] (National Institute for
Agricultural Research), 1997; Uaiene, 2004). Basically, soil nutrients are removed
through crop harvest and erosion, and farmers do not apply sufﬁcient quantities of
nutrients to build the soils to desired levels and to replace what is removed by the crops

and erosion.

ll

The most cultivated food crops in Mozambique are maize, cassava, beans and
groundnuts. Among these food crops, maize is the most grown food crop all over the
country. About 40.46 percent of cultivated land is planted to maize, and about 60 percent
is planted to cassava, beans, groundnuts, sorghum, rice, sweet potato, and millet.
Maize-based System

Maize (Zea mays L.) is the dominant food crop cultivated as an intercrop with
groundnuts, cowpea, and sorghum, pigeon peas, and/or pearl millet. There is mainly one
maize cr0p per year. Maize is sown in the rainy season from October to December and
harvested between April and May. Maize-based farming systems are heavily dependent
on rainfall, manual cultivation, and almost no use of purchased inputs such as improved
seed, inorganic fertilizers, and pesticides. For crop cultivation in general (which includes
maize cultivation), of about 3,064,715 agricultural enterprises, only 2.7 percent use
fertilizers, 3.9 percent use irrigation, 4.5 percent use pesticides, and 11 percent use animal
traction for crop cultivation (INE, 2001). The actual maize yields are much below the
estimated potential yield. The actual average maize yield is between 0.4 and 1.3 tons per
hectare; potential yields are between 5 and 6.5 tons per hectare (Jeje et al., 1998). Maize
yields can be increased if improved maize varieties and chemical fertilizers are accessible
to small farmers (Bias and Donovan, 2003).

The Uses of the Maize Crop

The main uses of the maize crop are home consumption and provision of income.
Commonly, households consume a reﬁned maize meal (Upswa or Chima) whose
processing requires numerous steps, including hand pounding, soaking, washing, drying,

and milling (Rubey, Ward, and Tschirley, 1997). Small proportions of maize produced in

12

Mozambique are used for animal feed (poultry and swine) (Fumo, 1995). Maize grain
provides farmers with monetary beneﬁts through farmers’ participation in the market.
For example, data from the agricultural survey for 1995-96 showed that, among the basic
food grains, only maize was commonly marketed, with about 20 percent of total
production sold and 21 percent of households participating in the markets for maize (Bias
and Donovan, 2003). Because the animal feed industry is an important factor of demand
for maize in Aﬁica (Byerlee and Heisey, 1997), its expansion in Mozambique may
increase farmers’ participation in the markets for maize.

Dissemination of Improved Maize T echnology

Despite the importance of the maize crop as a staple food and also as a source of
income, very few farmers use improved maize seed and chemical fertilizers. The result is
low maize yields, averaging between 0.4 and 1.3 t/ha (Jeje et al., 1998). Most farmers
use saved seed from their own production or seed offered by another farmer or relative
(Massingue et al., 2004).

To assist farmers in increasing maize production (and to some extent stabilize
maize yields), the public extension service, through the National Directorate of Rural
Extension (DNER), disseminates information on improved technologies in maize among
farmers in 66 of the 128 districts in Mozambique. The public extension service uses
demonstration plots, personal visits, group meetings, and radio broadcasts to educate
growers. Radio broadcasts have been particularly effective in the dissemination of
improved seed (Massingue et al., 2004). Recommendations for use of improved maize
seed emphasize seed treatment for the control of termites; a seed rate of 50,000 plants /

ha, which corresponds to 25 to 30 kg/ha of seed, proper seed spacing and planting depth

13

(DNER, 2002). Recommendations for enhancing soil fertility are of two types. First,
recommendations which emphasize low-input agriculture in which the soil resource base
is maintained through the use of organic compounds and other low-cost practices that
promote soil water and fertility conservation and avoid soil degradation through erosion.
Second, recommendations that promote agricultural intensiﬁcation, particularly through
the use of inorganic fertilizer, which is considered the key element for sustainable
agricultural production, productivity, and conservation of the natural resource base. An
integration of the two types of recommendations is seen as the most adequate approach to
fertilizer recommendations, mainly because organic matter is often insufﬁcient to fully
replenish and provide the quantities of various nutrients required by crops (Bias and
Donovan, 2003).
DNER/$02000 Extension Program

The National Directorate of Rural Extension and Sasakawa Global 2000 initiative
(DNER/SG2000 program) is an example of a national effort to promote agricultural
intensiﬁcation by the use of improved agricultural technology, primarily in maize. Since
1995, the DNER/SG2000 program has been disseminating improved maize varieties
(open-pollinated varieties and hybrids) and chemical fertilizers to improve maize yields
from 0.5 t/ha to the potential of 6 t/ha (DNER, 2002). These technologies were
disseminated by means of demonstration plots. Participating farmers managed one of
two types of demonstration plots: technology demonstration plots of 0.1 ha, for which
farmers managing this type of plots were given free inputs; and credit plots of 0.5 ha.
Farmers managing credit plots were delivered inputs on a credit basis, involving nearby

input suppliers and rural traders (DNER, 2002). At the end of the cropping season,

14

farmers had to pay back the loan with part of their production. In general, the program
was successful in encouraging farmers to use improved maize seeds and chemical
fertilizers (Gemo et al., 2005). The credit schemes did not always work well, however -
some farmers failed to pay back the loans (Jeje etal., 1998). It is important to note that
the credit component is a determinant factor in the use of improved maize seeds and
chemical fertilizers. The successful smallholder-led maize revolution in Zimbabwe
demonstrated this (Eicher and Kupfuma, 1997). Thus, if alternative credit schemes to
deal with delinquency rates are not in place in Mozambique, the future large scale-
adoption of improved technology may be threatened.

In the district of Manica, the DNER/SG2000 program was implemented in the
administrative posts of Machipanda and Vanduzi from 1996 to 2001 (personal
communication with head of Manica District Public Extension, 2005). From 1996 to
2001, DNER/SG2000 distributed an informational package on the new hybrid and
fertilizer technology, which included the improved maize variety Hybrid SC513 and
chemical fertilizers NPK 12-24-12 and urea (DNER, 2002). Seed and fertilizer

recommendations to attain 4 tons of maize grain per hectare are presented in Table 1.

Table 1. Seed and Fertilizer Levels to Produce 4 tons of Maize Grain per Hectare.

Technology Components Recommendations

 

Maize varieties SC513
Seed rate 50,000 plants/ha (25-30 kg of seed/ha)
Fertilizers 100 kg/ha of NPK (12-24-12) at planting time

200 kg/ha urea 30 and 60 days after emergence

Source: DNER, 2002.

15

Production and Distribution of Improved Maize Seed

In Mozambique, two private companies produce and import improved maize seed
from neighboring countries: the Mozambique Seed Company (SEMOC) and the
PANN AR, a South Africa-based company. Currently SEMOC is the largest private seed
company. Mozambique owns about 25 percent of SEMOC. The other 75 percent of the
company is owned by SEEDCO, a Zimbabwean Seed Company (personal
communication with general manager, 2006). The Hybrid maize SC513 is produced by
private farmers under contract with SEMOC. After the seed is produced, it is delivered
to SEMOC, where is processed, labeled, and sold. At the time of this research (March-
May 2006), the seed of SC513 could be bought directly from SEMOC for 24,000
MZM/kg (approximately $0.92 per kilogram [US $1= 26,000 MZM]).

The Hybrid maize variety SC513 is a white dent-type seed developed for high
yields (4 to 9 t/ha) and grey leaf spot tolerance. Other characteristics of Hybrid SC513

are presented in Table 2.

Table 2. Characteristics of Improved Maize Hybrid Variety SC513.

 

Characteristics Speciﬁcations

Manning rate Intermediate maturity: 137 days to maturity

Ear placement Relatively high

Grain drying rate Slow rate

Drought tolerance Moderate to good tolerance

Resistance to cob diseases Moderately resistant

Root lodging Slightly susceptive at high plant population density (over

44,000 plants/ha)

Source: Seed Company (SEEDCO), 2003-2004.

The commercial sector is in charge of providing improved maize seed to farmers
(Massingue et al., 2004). However, seed distribution is still weak in Mozambique,

because of an underdeveloped market, which leads to low demand for improved varieties

16

and low maize yields; and insufﬁcient coverage by the distribution network. The
distribution network does not cover all the districts, and some of the covered districts
have only one seed depot representing one of the two seed companies.

There are three ways by which improved seed reaches the farmers. The ﬁrst
distribution arrangement involves the formal commercial (private) sector. Here, the
wholesalers/retailers acquire seed (on credit a basis) from the seed companies. The
wholesalers/retailers may or may not represent the seed company. Those that represent a
seed company usually have the name of the company written on the wall of the shops.
The wholesalers/retailers’ main task is to provide seed to retailers in the area. Both the
wholesalers/retailers and the retailers can sell seed to the farmers. At the time of this
research (March-May, 2006), farmers reported a price between 25,000 and 30,000
MZM/kg.

The second distribution arrangement involves the public sector. The government
buys large quantities of seed from the seed companies for purposes of emergency
programs. The seed is then distributed to the farmers for free. Extension programs such
as DNER/SG2000 also provided the participating farmers with improved seed.

The third distribution arrangement involves the informal commercial sector.
Among these three seed providers, the public sector and the informal traders have the
major market shares. The basic distribution channel (excluding seed exchange within the

community) is summarized in Figure 1.

17

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wholesaler Retailer
/Retailer
CoSrreregn Emergency Farmer
p y /Extension \j‘
Informal
Traders

 

 

 

 

 

 

 

 

 

Figure 1. Distribution Channels for Improved Maize Seeds in Mozambique.

Production and Distribution of Chemical Fertilizer

Unlike improved maize seed, nitrogen and phosphorus fertilizers are not currently
produced in Mozambique. The country has one of the lowest use rates of fertilizer in
sub-Saharan Aﬁica, with an estimated average annual application of 1.84 kg nitrogen-
phosphorus-potassium (N PK) fertilizer per hectare, compared with an average of 16.55
kg/ha in West Africa and 8.89 kg/ha in all sub-Saharan countries (Bias and Donovan,
2003). Fertilizer imports are very low. For example, between 1983 and 1997,
Mozambique imported between 1,100 and 7,269 tons of nitrogen and phosphorus
fertilizers (DNER/SG2000, 2003), and the country is still importing less than 10,000 tons
of fertilizer per year (Bias and Donovan, 2003).

Jeje at al. (1998) describe three ways used by large agricultural enterprises to
acquire agrochemicals. First, the large agricultural enterprises could order them through
agrochemical companies representing multinational ﬁrms (e. g., Agroquimicos, Tecap,
Zeneca). These companies supplied pesticides worth an estimated $3 million to $5

million per year and met approximately one-third of total demand. The second option

18

involves joint venture companies and other large commercial enterprises ordering
pesticides directly. This accounts for another third of total pesticides used. Thirdly, the
Japanese KRII aid program supplied the rest of the national pesticide demand and
virtually all of the fertilizer used in Mozambique (Jeje et al., 1998).

Fertilizers are not accessible for small farmers. According to farmers’ reports, the
price of 100 kilograms of NPK was about 1,200,000 MZM (US $46.20) and the cost of
50 kilograms of urea was about 445,000 MZM (US $17.10), at the time of this research
(March-May 2006). Most of the imported fertilizer is used by large farmers for the
production of cash crops, mainly tobacco. One way that small farmers obtain fertilizer is
through production of cash crops such as tobacco under contract with larger farmers.
Small farmers involved in these contracts are provided with fertilizer. There is little use
of fertilizer on maize, and such use would be risky given crops’ price variability and
market instability, as well as climatic risk (Bias and Donovan, 2003). The basic

distribution channel for fertilizers is summarized in Figure 2.

 

Contract growing schemes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Agrochemical Large
companies agricultural
representing enterprises
multinational ﬁrms
Fertilizer ' Farmer
producers
outside I 4 l
Mozam-
bique ,
Extensron
Japanese KRII Government
' aid program <
Retailer r

 

 

 

 

 

 

 

 

 

Figure 2. Distribution Channels for Chemical Fertilizers in Mozambique.

19

Proﬁtability of Improved Maize Technology

Jeje et al. (1998) assessed the proﬁtability of maize production technology
disseminated by the DNER/802000 program in 1996 and 1997 in the Manica province.
Their results suggest an association between proﬁtability and the period in which farmers
sell maize. The study revealed that when farmers sold maize in June, only 36 percent
made a proﬁt. At the December price, 80 percent made a proﬁt. When farmers sold
maize midway between July and December, 62 percent proﬁted. Thus, there are
indications that some farmers do make a proﬁt when they use improved maize varieties
and agrochemicals. It is important to note that the proﬁtability study by Jeje et a1. (1998)
was conducted with higher levels of inputs than those used currentely (Appendix B). For
this reason, the results may or may not apply to the current farming situation in
Machipanda and Vanduzi. A similar study to that by J eje et al. would have to be
replicated to ﬁnd out whether some farmers are still making money with the current
levels of inputs.

Though proﬁtability asssement was not the objective of the current study, it did
perform rough estimations on returns to farniliy land and labor with Hybrid maize SC513
and chemical fertilizers NPK and urea, excluding the levels of pesticides applied as
described in the standard technology recommendation (Appendix B). The maximum
average maize yield of 1.3 ton/ha was assumed as the actual maize yield with local
maize, and the minimal potential average maize yield of 4 ton/ha was assumed as the
actual yield with SC513 and chemical fertilizers. Table 3 presents the average prices for

inputs and maize grain.

20

Table 3. Average Prices of Inputs and Maize Grain.

 

Item Price (MZM) Price (US $)
Improved maize seed 27,500 MZM per kilogram 1.10
NPK(12-24-12) 1,200,000 MZM per 100 kilograms 46.20
Urea 445,000 MZM per 50 kilograms 17.10
Local seed 3, 700* MZM r kilo ram 0.10

Source: Fanners’reports and TIA, 2005*.

Tables 4, 5 and 6 present the budgets and partial budgeting procedure to estimate

the net returns to family land and labor.

Table 4. Budget for Hybrid Maize SC 513.

 

Income Sources Quantity per ha Unit Price per unit Total per ha
Grain (maize) 4 ton 3,700,000 14,800,000
Total Income 14,800,000
Expenses

Seed 25 Kg 27,500 687,500
NPK (12-24-12) 2 50 kg 600,000 1,200,000
urea 4 50 kg 445,000 1,780,000
Sub-total 3,667,500
Interest on current debt 293,400
(8%)

Total ExEnses 3,960,900

Table 5. Budget for Local Maize.

 

Income Sources Quantity per ha Unit Price per unit Total per ha
Grain (maize) 1.3 ton 3,700,000 4,810,000
Total Income 4,810,000
Expenses

Seed 40 kg 27,500 1,100,000
Sub-total 1 ,1 00,000
Interest on current debt 88,000
(3%)

Total Expenses 1,188,000

21

Table 6. Partial Budgeting: Change from a Local to Hybrid Maize.

 

Reduction to Income Gains to Income

(MZM/ha) (MZM/ha)

Added Cost 3,960,900 Added revenues 14,800,000
Reduced Revenues 4,810,000 Decreased cost 1,188,000
Total Reductions (A) 8,770,900 Total Gains (B) 15,988,000
Net Change in income (B-A) 7,217,100

The net gain of growing improved maize in (2005-06) was positive 7,217,100
MZM (US $277.58). This value would decrease slightly if costs of pesticides were
included, but farmers would still make some money with 4 tons of grain maize SC513 per
hectare. Therefore, it would be favorable for the family farm to adopt the Hybrid maize
SC513, provided that farmers apply the technology as recommended, are risk takers, and

prioritize the objective of making proﬁt.

Section 2. Beneﬁts and Disadvantages of Using Improved Maize and Chemical

Fertilizer

For rain-fed farming (as is the case in Mozambique), improved maize varieties
and chemical fertilizers provide farmers with alternatives to respond to risks and
opportunities associated with environmental and socioeconomic factors.

Environmentally induced Risks to Maize Production and Fertilizer Use

In their study “Yield Response to Fertilizer Application in Kenya,” Rotter and van
Keulen (1997) identiﬁed the following weather-induced risks to maize production and
fertilizer use: crop failure due to temporary but distinct dry spells and/or soil water
deﬁcits due to low seasonal rainfall and shallow soils; considerable yield depression due

to soil and runoff water loss (this includes decreases in soil quality due to soil

22

degradation after heavy showers); yield reduction due to pests and diseases; and
considerable yield limitation due to a temporary excess of water.

Environmentally induced risks, particularly those associated with low, erratic, and
unreliable rainfall, were the main risk factors in the semiarid areas (Rotter and van
Keulen, 1997).

Socioeconomic Induced Risks to Maize Production and Fertilizer Use

Socioeconomic risk factors include the possibility of ﬁnancial loss due to low
maize prices and increased costs of mineral fertilizer. Rotter and van Keulen (1997)
found that socioeconomic factors were more relevant in semihumid areas. In general,
fertilizer is the most costly cash input used by the typical smallholder in southern Africa
(Kumwenda et al., 1997). Byerlee (1994) mentioned the high price of chemical fertilizer
as a constraint to fertilizer application among maize growers in southern Aﬁica. While
studying perceptions and attitudes toward soil fertility technologies in western Africa,
Enyong, Debrah, and Bationo (1999) found that farmers perceive urea as being more
costly than NPK because it had less residual effect and must be purchased yearly.
Possible .Adaptations to Prevailing Risks

To adapt to variables in environmental conditions and associated risks, farmers
can choose to cultivate improved maize varieties and use chemical fertilizer. Researchers
have developed maize varieties that are tolerant of drought and low nitrogen levels,
resistant to ﬁeld diseases (including maize streak virus disease and grey leaf spot),
resistant to attacks from stem borers, lodging and herbicide injury, and have enhanced

seed vigor and germination. These maize traits improve and stabilize yield.

23

Maize varieties have also been developed with resistance to storage pests. This
contributes to reduction in postharvest losses. Vowotor, Bosque-Perez, and Ayertey
(1994), and Meikle et a1. (1997) found that the use of improved ﬂint maize varieties,
combined with the practice of storing unshelled maize increased the development period
of the weevil and reduced the possibility of the build up of destructive populations of
Sitophilus zeamais. When farmers use maize varieties resistant to storage pests, they are
able to store maize and sell it when prices are relatively high, increasing the chances of
making a proﬁt.

Improved maize varieties can help farmers face labor constraints, food insecurity,
and lack of income. Research in southern Africa (Byerlee, 1994; Kaliba et al., 2000)
indicated that farmers prefer early-maturing maize to better deal with labor constraints,
risk considerations and crop rotations. Sometimes farm households will improve food
security by planting an early-maturing variety that can be consumed in the “hungry
season” before the main harvest. Income, either for investments or buying food, is the
measure of success for farmers (Rotter and van Keulen, 1997). Early-maturing varieties
provide sources of cash for farmers in areas with small-scale irrigation and nearby market
places for fresh maize (roasting ears).

Farmers can also beneﬁt from applying chemical fertilizer to maize. When used
in optimum amounts, chemical fertilizers increase production and farm efﬁciency
(Rogers and Havens, 1961). In Kenya, Rotter and van Keulen (1997) found that there
was a tremendous potential for increasing maize yields, and hence national production,

by applying moderate amounts of nitrogen (N) and phosphorus (P) fertilizers in the

24

lowlands and midlands. Treatments with dry mix NPK brought economic beneﬁts in

terms of grain yield returns (Belay, Claasens, and Wehner, 2002).

Section 3. Deﬁnitions and Theoretical Views about Attitudes

Several researchers view “attitude” as a learned predisposition to respond in an
evaluative manner toward an “attitude object” - i.e., situations, institutions, concepts or
persons. Oskamp and Schultz (2005), after reviewing several deﬁnitions of “attitudes”,
deﬁned attitude as “a predisposition to respond in favorable or unfavorable manner with
respect to a given attitude object” (p. 9). Likewise, Oskamp and Schultz (2005), and
Eagly and Chaiken (1993) emphasize attitudes as evaluative responses. Eagly and
Chaiken (1993) deﬁne attitude as “an evaluative state that intervenes between certain
classes of stimuli and certain classes of evaluative responses, which express approval or
disapproval, favor or disfavor, liking or disliking, approach or avoidance, attraction or
aversion toward the attitude object” (p. 3). According to Eagly and Chaiken (1993),
evaluative responses are provided for verbal statements of beliefs, verbal statements of
affect, and verbal statements concerning behavior toward the attitude object.

Eagly and Chaiken’s view of attitudes is similar to the latent process viewpoint of
Deﬂeur and Westie (1963) referenced by Oskamp and Schultz (2005). Under the latent
process viewpoint, attitude is viewed as a process occurring within the individual.
Attitude is used to explain the relationship between stimulus events and the individual’s
responses. In this sense, an attitude is an intervening variable. Attitude is a theoretical
construct that it is not observable in itself but mediates the relationship between certain

observable stimulus events (the environmental situation) and certain responses.

25

Concerning the class of evaluative responses, cognitive responses refer to the
thoughts or ideas that people have about the attitude object. These thoughts, cognitions,
knowledge, or opinions are often conceptualized as beliefs. A belief is a statement that
expresses the relationship between attitude object and characteristics of the attitude
object. The aﬂective responses consist of feelings, moods, motives, emotions, and
associated psychological changes and verbal statements of affect in relation to the attitude
object. The behavioral responses cover people’s covert actions and both intended and
actual verbal statements concerning behavior.

This study adopts the tripartite view of attitudes and the conceptualization of
attitude provided by Oskamp and Schultz (2005), which states, “An attitude might be
conceptualized as a summary of all of a person ’s evaluative beliefs about, affective
reactions toward, and behavioral responses to an attitude object” (p. 14). Although the
deﬁnition contains three types of responses, they are not always simultaneously present

in every attitude (Eagly and Chaiken, 1993).

26

.333 now—menu use Ewem use @0ch Nag—Low use mgmo 88m 3:55. .movﬁuﬁx Mo “50955 $8on Eoﬁq 2:. .m oBmE

 

 

 

.zozoE

new 25:0 883 a m_ 8:03.05

SEE com EonEtow E3823 we 83

5258 use 2:: me 833 e a

being 059: 338:: .«o gags—=0
8:38.. ESSEX—

 

_ 3368.3 3.532.3—

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

06 com poem no: ﬂ bum—Ea _8_Eono M” n. bumtﬁﬁomsmf
.voow mouwﬁ Hm. EB EKQER Rka _ 3332A. o>=oet< m mug—Eu.“ 33:55 wad
oﬂeﬁ 335:: Lo 92:: qutom 283:3?» Ago—Eek w 1 8:03; oﬂaE @9895
8:33.. 953...: m ovsxtmw m we mocmtuﬁmﬁno
. .0382?» mm Suite.“ 32525 _ 8839:. 95:.on
Eﬁﬂmoznwaohu m_ 368 335::
omuoamo. e>§=ue0
82.2.8.— o>=«=_a>m 8:0er wing-31—
$33.5.»

833E?» «no—Eamon. oEanEO 033.3.» “£5285 8339:— 2531— 2363935 035E039

£3455.“ Eugene

can monomer» 3E8 vegan: 8 BER £8358 09:33 pee momentum mo sesame—u 08 go “893%? an 959% m 833m

27

As shown in Figure 3, the stimulus events can arouse a person’s cognitive,
affective, and/or behavioral processes, many of which are non-observable. These
processes (together or separately) can produce an attitude toward objects involved in the
stimulus situation, and this attitude is a non-observable variable or latent construct. The
presence of an attitude will give rise to, and can be measured in terms of, observable
responses, which may be cognitive, affective and/or behavioral reactions to the attitude
object (Oskamp and Schultz, 2005).

The Formation and Change of A ttitudes

 

Experimental research has demonstrated that attitudes are formed and modiﬁed as
people gain information about the attitude object (Eagly and Chaiken, 1993). In the
tripartite view of attitude, attitudes are formed through cognitive, affective and behavioral
learning processes.

Cognitive learning processes take place when the individual, through direct and
indirect experience, gains information about the attitude object and then forms beliefs. In
aﬂ'ective learning processes, attitudes are formed on the basis of affective or emotional
experiences. Attitude is a product of pairing an attitude object with a stimulus that elicits
an affective response. In behavioral learning processes, attitudes derive from past
behavior. Individuals take into account the conditions under which previous behavior
was formed to infer attitudes that are consistent with their prior behavior (Eagly and
Chaiken, 1993).

Other learning processes include persuasion, instrumental conditioning learning,
and observational learning (Oskamp and Schultz, 2005). Persuasion is an interactive

process through which an individual’s knowledge and beliefs are changed by means of

28

communication (Murphy and Alexander, 2004). Instrumental conditioning learning
occurs when the individual’s behavior is the means by which reward or punishment is
achieved (Oskamp and Schultz, 2005). Observational learning is the common type of
learning. It occurs, without any external reinforcement, through imitation of others’
behavior.

Attitude and Behavior Relationship

Oskamp and Schultz (2005) argue that “. . .as an intervening variable ‘attitude’ is a
useful concept only if it conveniently summarizes, or predicts, or is related to patterns of
actual behavior” (p. 266). This argument suggests that social researchers would produce
more valuable results if, in their studies on attitudes, they include analysis of attitude-
behavior (A-B) relationships.

Attitude-behavior (A-B) relationships are important for predicting the general
tendency that the individual will engage in behaviors relevant to the attitude object
(Eagly and Chaiken, 1993). From the point of view of extension educators, it is
important that attitudes serve a predictive purpose if measurements of support for using
improved maize production practices are useful tools in guiding program planning.

In Mozambique, agricultural extension agencies implement programs to assist
farmers in improving maize productivity through the use of improved maize seed
varieties and chemical fertilizers. The success of the extension programs depends on
farmers’ volitional behavior - i.e., farmers decide whether to perform behaviors related to
the use of improved varieties, such as buying the seed, planting the seed, processing corn
for animal feed, cooking corn for consumption, and behaviors related to the use of

chemical fertilizers, such the application of NPK and urea on maize.

29

Extension ofﬁcers can assess farmers’ attitudes toward improved maize
production practices and generate measurements of support to improved maize
technology as part of monitoring program implementation. For these measurements to be
useful tools, they must serve a predictive purpose with respect to farmers’ behavior.

In many cases, attitudes and actions are quite different (Rogers, 1995), but it is
essential to explore the association between attitudes and behaviors (A-B) and formulate
hypotheses on the predictability of farmers’ behavior from attitudes. For example, Kim
and Hunter (1993) suggest that A-B correlations do exist and can take low (.26),
moderate (.50), and high values (.86). These researchers also make it clear that a
conclusion about the existence or non-existence of an A-B relationship is valid only if the
measures of attitudes and behaviors are conceptually similar.

Factors Affecting Attitude-Behavior Relationship

According to Oskamp and Schultz (2005), the relationship between attitude and
behavior is affected by the following factors:

Attitude certainty. This is the degree to which cognitive and affective components are in
agreement. Attitude certainty is low when people have ambivalent attitudes - that is, both
high positive and high negative attitudes toward an object.

Instability of attitudes. The stability of attitudes decreases over time; therefore, the
longer the gap between the time when attitudes are measured and the time when the
behavior is measured, the greater the chance that attitude-behavior relationship will be
inconsistent.

Direct experience. Attitudes learned through direct experience with the attitude object

are the most predictive of behavior.

30

Attitude accessibility. Accessible attitudes tend to best predict behavior (Oskamp and
Shultz, 2005). Accessible attitudes tend to be those learned through vivid information,
such as results demonstration plots. Accessible attitudes are easily retrieved from the
memory. Accessibility of information provided via concrete examples is explained by
the principle that information is weighted in proportion to its vividness (N isbett and Ross,
1980)

Behavior. Severity of the expected consequences of the behavior can decrease attitude-
behavior correlations.

Situation. The inﬂuence of situation occurs when something not thought about happens
that forces the individual not to behave in accordance with his/her attitude.

Person. Low self-monitoring individuals (i.e., individuals who do not change their

behavior according to the situation) tend to show high attitude-behavior correlations.

Section 4. Empirical Studies on Farmers ’ Attitudes Toward Agricultural T echnologv

Studies on farmers’ attitudes toward agricultural technology have examined
causal relationships between attitudes and farmers’ behavior; associations between
attitude and individual characteristics such as age, years of experience, knowledge; and
associations between attitude and information sources. Ganpat and Bholasingh (1999)
examined attitudes toward technology implementation and attitudes toward farming as a
challenge among Trinidad’s farmers. The researchers found that younger farmers had
stronger beliefs about technology and felt that farming was more of a challenge than
older farmers.

Rogers and Havens (1961) utilized an experimental design to measure the impact

of demonstrations on farmers’ attitudes toward fertilizer. The results showed that more

31

educated farmers had a positive attitude change toward fertilizer usage. The knowledge
of fertilizer (i.e., how to use fertilizer, what nutrients the crop needs) acted as an
intervening variable between “attitude toward fertilizer” and “use of fertilizer.” Soil tests
were mentioned by farmers as the most important sources of information on the use of
fertilizer, while farmers’ own experience and demonstrations occupied the second and
third positions (Rogers and Havens, 1961).

A study by Chilonda and Van Huylenbroeck (2001) was based on the assumption
that differences in attitudes and socioeconomic circumstances inﬂuence the uptake of
veterinary services by small-scale cattle farmers. It employed a logit model to estimate
the probability of use of veterinary services in the eastern province of Zambia. The
empirical logit estimates from a sample size of 209 farmers indicated that a positive
attitude toward utilization of veterinary services favored the use of these services.
Farmers who had a positive attitude toward utilization of veterinary services had more
experience in cattle keeping, and knowledge of husbandry practices, and enjoyed
additional earnings from hiring of oxen for draﬁ power (Chilonda and Van
Huylenbroeck, 2001).

Willock et a1. (1999) studied the differences in attitudes toward chemical use between
business-oriented farmers and environmentally oriented farmers in Edinburgh, Scotland.
The results showed that attitudes toward chemical use inﬂuenced only environmentally
oriented farmers. Signiﬁcant association (P<.01) of about 0.2 was found between
environmentally oriented behavior and attitudes toward chemical use (Willock et al.,

1999)

32

Lichtenberg and Zimmerman (1999) examined the inﬂuence of information from
various sources on farmers’ attitudes about the effects of pesticide use on human health
and the environment. Regression analysis showed that farmers who placed greater
importance on information from the news media expressed greater concern about all
forms of environmental quality problems associated with agrochemicals, while farmers
who placed greater importance on cooperative extension information expressed greater
concern about speciﬁc aspects such as mixing and loading pesticides, and residues on
food and in drinking water.

There are also studies on attitudes (not necessarily within agriculture) that are
informative about the relationship between attitudes and age. Some of these studies
found that older respondents were more prone to express negative attitudes (Wang,
Lassoie and Curtis, 2006). Diallo (1983) referenced McGuire (1969) to explain that
positive attitudes decline with age.

Other studies on farmers’ attitudes toward agricultural technology did not look
much at the relationships between attitudes and other individual characteristics. Rather,
the studies provided a description of the attitudes held by the farmers. Diallo (1983)
interviewed 30 and surveyed 109 maize growers to determine farmers’ attitudes and
experiences with no-till corn production systems in Baltimore County, Washington, in
the United States. The ﬁndings indicated that farmers tended to hold positive attitudes
toward conservation tillage. Thompson (1992) conducted interviews with 35 rice
growers in Bangladesh to explore their perceptions and attitudes about the use of urea

deep placement (UDP) technology as a mean to reduce nitrogen use and improve

33

fertilizer efﬁciency. The study reports that farmers had very favorable attitudes about
UDP and perceived the practice as beneﬁcial for increasing rice production.

Positive attitudes toward fertilizer use were also found through qualitative studies.
Enyong et al. (1999) used group discussions, interviews and ﬁeld visits to investigate
subjective and cultural processes underlying western African farmers’ practices and
attitudes toward adoption of NPK and urea. The data from a non-random sample of 117
farmers indicated that most farmers recognized the beneﬁts of chemical fertilizer for
enhancing soil fertility. The study also identiﬁed reasons for adoption of fertilizer use
(which included easy use or application of chemical fertilizers, immediate and visible
effect during the crop season, and residual effects for subsequent seasons) and non-
adoption (which included non—availability or scarcity in market, lack of access to credit to

purchase fertilizer, and under application).

Section 5. Theoretical Measurement Models for Development of A ttitudinal Scales

This section uses the reviewed literature to formulate hypotheses about theoretical
measurement models that describe the relationship between the possible underlying
factors (attitudes) and the observed variables (items) (Knoke, Bonrnstedt, and Mee,
2002)

The reviewed literature on beneﬁts and disadvantages of improved maize varieties
and chemical fertilizers suggests that farmers operate in an environment of risks
associated with climate (drought, soil, and runoff water loss, pests, and diseases) and
socioeconomic factors (ﬁnancial losses from low maize prices and increased costs of
mineral fertilizer). The literature on attitude formation implies that maize growers

exposed to information about climatic factors develop beliefs about improved maize

34

varieties and the effects of chemical fertilizers to cope with environmental hazards. In
addition, the literature on attitude formation implies that maize growers exposed to
information about socioeconomic risk factors develop beliefs about the increased income
and lower costs incurred by adopting improved maize varieties and chemical fertilizers.

Attitudes formed about use of improved maize varieties suggest the presence of
two factors: production characteristics and income factors. Similarly, for chemical
fertilizers, the literature suggests the presence of two factors: eﬂect of fertilizer and cost
factors. The statements (items) used to capture each factor are presented below.
Attitudes Toward Characteristics of Improved Hybrid Maize SC 5 1 3

- When rainfall is low, local maize variety Chimanhica produces better than
improved maize variety SC513.

- When rainfall is low, local maize variety Candjere produces better than improved
maize variety SC513.

- With adequate rainfall, it is better to plant improved maize SC513 because it
produces better than local maize varieties.

- Porridge made of improved maize variety SC513 tastes good.

- Grain of improved maize variety SC513 is good for hand pounding.

- Grain of improved maize variety SC513 is good for milling.

- Improved maize variety SC513 has good germination.

- Improved maize variety SC513 is more resistant to storage pests than local maize
Candjere.

- Improved maize variety SC513 is more resistant to storage pests than local maize
Chimanhica.

- Farmers in this area should plant improved maize variety SC513.

Attitudes Toward Income from Improved Hybrid Maize SC 5 I 3

- Planting improved maize variety SC513 provides more income than local variety
Chimanhica.

- Planting SC513 provides more income than local variety Candjere.

- Grain of SC513 is easy to sell.

- Fresh maize (roasting cobs) of SC513 is easy to sell.

- Planting SC513 is waste of time and money.

- Seed SC513 is easy to ﬁnd.

35

Attitudes Toward Effects ﬁNPK and Urea

- NPK/urea is good for the maize crop.

- NPK/urea increases maize yield.

- NPK/urea is not good for the soil.

Farmers in this area should apply NPK/urea on maize.

Attitudes Toward C osts of NPK and Urea
- NPK/urea is expensive.
- NPK/urea is a waste of time and money.
- NPK/urea is risky because of crop failure due to weather conditions.
- NPK/urea is easy to obtain.

It is reasonable to expect associations between the attitudinal dimensions
(factors). Farmers may appreciate the production characteristics of the improved maize
varieties and be positive about the improved maize variety output as a marketable good.
In Honduras, for example, farmers who planted hybrid maize were focused on
commercialization and appreciated the production characteristics of the hybrid variety
(Hintze et al., 2003). Farmers may also appreciate the effect fertilizer has on maize yield
but be put off by the high costs of fertilizers.

On the basis of the assumptions presented above, the following is hypothesized
about the attitudinal dimensions (factors):

H1: Attitudes toward production characteristics of SC513 are positively associated
with attitude toward income from SC513.

H2: Attitudes toward the effect of NPK and urea on maize yield are negatively

associated with attitudes toward costs of NPK and urea.

36

Section 6. Adoption and Diffusion T heory

Individual or F arm-level Adoption

According to Feder, Just, and Zilberman (1985), individual or farm-level adoption
is deﬁned as the degree of use of a new technology in long-run equilibrium, when
farmers have complete information about the new technology and its potential. Hence,
adoption occurs when farmers decide to use the new technology on basis of information
given (Kaliba, Verkuijl, and Mwangi, 2000). For divisible technologies such as
improved maize varieties and chemical fertilizers, the degree of adoption is measured by
the area the farmer dedicates to utilizing the technology (Feder et al., 1985). It is
important to note that there is no deﬁnition of adoption that applies to all situations. For
each situation the researcher deﬁnes the criterion of adoption. The criterion may consider
farmers who plant even a few rows of the new variety as adopters (CIMMYT, 1993).
This is the criterion used in this study.

Individual (farm-level) adoption studies try to explain the factors that inﬂuence
farmers’ decisions about whether to use the new technology (Hintze et al., 2002).
Researchers use various analytical procedures for studying innovation adoption by

farmers.

Analytical Procedure Focusing on Measurement of Variables Inﬂuencing Adoption

For example, considerable research (Feder et al., 1985; Kaliba et al., 2000;
Abebaw and Belay, 2001; Tesfaye and Alemu, 2001; Tesfaye, Bedassa, and Shiferaw,
2001) has been conducted to determine factors associated with adoption of improved

agricultural technology. The analytical framework used by these researchers consists of

37

descriptions of the regression equation (in most of the cases, logistic regression),
descriptions of variables inﬂuencing adoption, and formulation of hypotheses about the
associations between variables and adoption. The variables hypothesized to inﬂuence
adoption include level of education, household size, use of hired labor, experience of
farmers, access to credit, access to extension services, beliefs, human capital, weather,
soils, and availability of water. F eder et a1. (1985) conceptualize adoption as a
quantitative variable and differentiate themselves from researchers who view adoption as
a process involving mental stages that range from ﬁrst hearing about an innovation to

ﬁnal adoption.

Analytical Procedure Focusing on Mental Processes and Overt Behavior

Rogers (1995) conceptualized the innovation-decision process to explain how
farmers deal with uncertainty associated with new technology. Hence, the stages
involved in the innovation-decision process can be thought of as factors that inﬂuence
farmers’ decisions about whether to use the new technology. The innovation-decision
process involves series of actions and choices over time through which an individual
evaluates a new idea and decides whether to incorporate the innovation into on-going
practice. The actions and choices in the innovation-decision process are grouped in ﬁve
stages: knowledge, persuasion, decision, implementation, and conﬁrmation. These
stages can be further subdivided into stages that involve strictly mental exercise and
stages that involve overt behavior (Rogers, 1995).

Stages Involving Mental Exercise

Knowledge stage. The knowledge stage occurs when the individual becomes

aware of the technology and gains understanding of the workings of the technology

38

(Rogers, 1995). Rogers differentiates between three types of knowledge. Awareness
knowledge refers to awareness of the existence of the technology. Awareness knowledge
can be gained by hearing about the technology and/or having seen the technology. The
second type of knowledge is how-to knowledge. How-to knowledge is motivated by
awareness knowledge and consists of information necessary to use the technology
properly. Examples of how-to knowledge include knowledge of the amount of fertilizer
to apply on one hectare of crop, the method, and the timing for fertilizer applications.
According to Rogers (1995), an inadequate level of how-to knowledge is likely to lead to
rejection and discontinuation of technology. The third type of knowledge is the principle
knowledge. Like how-to knowledge, principle knowledge is motivated by awareness
knowledge. Knowing why topdressing with urea should be applied when the maize is
knee-high is an example of principle knowledge. Knowing that plants require nutrients to
grow is another example.

Persuasion stage. The persuasion stage occurs when the individual, after being
exposed to the technology and gained knowledge about it, forms positive or negative
attitudes toward the technology. The attitudes will predispose the individual to accept
(adopt/use) or reject the technology. Adopters tend to hold positive attitudes toward the
technology (Rogers, 1995). According to Rogers (1995) at the persuasion stage,
messages from speciﬁc information sources, such as personal communication, are more
effective than mass media messages. Consistent with this, Valente’s (1993) study on
diffusion of innovation and policy decision making showed that diffusion of hybrid corn
occurred via interpersonal inﬂuence. Ryan and Gross (1943) explained that neighbors

were cited more frequently by farmers as the most inﬂuential (persuasive) source of

39

information on Hybrid corn, while the market (salesmen, commercial representatives)
was credited with informing the majority of operators. Thus, though salesmen are very
important as sources of introductory knowledge about the innovation, neighbors, through
their experience within the community, are expected to inﬂuence more in terms of action
(Ryan and Gross, 1943). Suvedi et al. (2000), through a study on farmers’ perspectives
of Michigan State University Extension (1996-1999), also found that farmers were
interested in one-on-one interaction with extension agents.

The information source can be associated with the level of speciﬁcity of attitudes.
Research (Lichtenberg and Zimmerman, 1999) suggests that extension inﬂuences farmers
to form more speciﬁc attitudes (such as the effect of pesticides on food) in contrast to
more general attitudes facilitated by the news media (such as attitudes toward all forms of
environmental quality problems).

Decision stage: At the decision stage, the individual engages in activities that
lead to the choice to adopt or reject an innovation (Rogers, 1995). According to Rogers,
adoption means that the individual has decided to make full use of an innovation as the
best course of action available.

Overt Behavior

Implementation stage. At the implementation stage, the individual puts the
technology into use (Rogers, 1995). Farmers may choose to adopt one or more of the
complementary components rather than adopting the technological package all at once to
reduce uncertainty about the success or failure of the new technology (Kaliba et al., 2000;
Byerlee and de Polanco, 1986). Moreover, farmers may decide to adopt the major

technical innovation from the package (e. g., improved maize seeds) and choose to adopt

40

other complementary components (e. g., fertilizer or pesticides) as they learn by doing
(Kaliba et al., 2000). Farmers follow a “stepwise adoption” characterized by a sequential
pattern, such as variety-fertilizer-herbicide, which is determined by estimated returns of
each component in a speciﬁc production zone (wet or dry) (Byerlee and de Polanco,
1986). This implies that adoption studies must ask speciﬁc questions about each
component of the package and take into account that individual components may be
adopted at different times and under different conditions (CIMMYT, 1993).
Confirmation stage. At this stage, the individual seeks information that supports
his/her decision (Rogers, 1995). If the individual has decided to adopt the innovation and
at the conﬁrmation stage he/she receives negative feedback about the innovation,
discontinuation may occur. Other reasons for discontinuation include replacement,

dissatisfaction, and misuse of the innovation.

Analytical Procedure Combining Measurement with Innovation-Decision Process

Some adoption studies use analytical frameworks that combine analytical
procedures focusing on measurement of variables inﬂuencing adoption with some
elements of the innovation-decision process. For example, Chilonda and Van
Huylenbroeck (2001) used a logistic regression model to determine the inﬂuence of
attitudes, years of experience, and knowledge of the use of veterinary services. The use
of logistic regression is characteristic of an analytical procedure focusing on
measurement of variables inﬂuencing adoption, while “attitudes” and “knowledge” are
variables described in the innovation-decision process. Neupane et a1. (2002) also used a
logistic regression model to determine factors associated with adoption of agroforestry in

the hills of Nepal. These researchers selected variables affecting adoption of forestry

41

technologies from an analytical framework that combined socioeconomic variables with
some elements of the innovation-decision process. The analytical framework is presented

in Figure 4.

 

Community characteristics
Natural environments, extension, employment opportunities, access to
market infrastructure, technology, education, local/indigenous knowledge

v

Household characteristics
Socioeconomic: Family size, age, gender roles, education, ethnicity, and
migration pattern
Resources: Labor force, land holding, ownership, livestock holdings
Others: Extension contacts, membership in local NGOs/farmers’ groups and

 

 

 

 

 

 

 

 

 

 
 
    
 

 
    
  

Attitude
toward
technology

    
   

 
   
  
 
 

Adoption
of technology

Awareness
of technology

 

4

 

Local NGOs/meers’ Groups
Coordination, local level participation, awareness campaign, meetings, local
resource mobilization, and moral support

4

Extem_al Extension Organization
Design and dissemination of appropriate maize technology
Design of farmers’ cross /study visit, demonstration farms and on farm trials

 

 

 

 

 

 

 

 

 

Figure 4. Technology Adoption Framework (N eupane et. al., 2002).

The framework presented in Figure 4 is widely applied to investigate the adoption
pattern of high-yielding cereal varieties and related practices. This framework portrays
many of the variables measured through analytical procedure focusing on measurement

of variables inﬂuencing adoption. The framework also shows awareness of technology

42

and attitudes toward technology, two variables described in the innovation-decision
process.
Aggregate Adoption

While adoption of technology by individual farmers at the current time provides
information on current practices, the diffusion process helps to document and describe
the pattern of adoption (CIMMYT, 1993) at the aggregate level on the basis of
information on past practices. Feder et a1. (1985) deﬁne aggregate adoption as the level
of use of a speciﬁc new technology within a given geographical area or population.
Adoption studies at the aggregate level focus on comparisons across geographic areas and
use the proportion of farmers employing the new technology in different regions (Hintze
et al., 2002).

In the innovation-decision tradition, aggregate adoption is conceptualized as a
result of the diffusion process whereby an innovation is communicated over time among
the members of a social system (Rogers, 1995). This process comprises four elements:
the innovation (technology), communication channels (sources of information), time
(year or month), and the social system (interrelated individuals or organizations).

The diffusion process is represented by the S-shaped curve Time versus
Cumulative Percent of Adoption (Figure 5). As explained by Rogers (1995), “when the
number of individuals adopting a new idea is plotted on a cumulative frequency basis
over time, the resulting distribution is an S-Shaped curve” (p. 22-23). This curve
describes the rate of adoption - i.e., the percentage of adopters in a given period (a year or

month, for example).

43

63:0 Bangui on“ 3 BEBEQQM $395 ".2955 2:. .m Sawmm
.Eo.:c>o==~\o =23§Q .2 mm 635 meowom Hacksaw

 

 

 

8o—
*8
$8
*3
oven
*8

norrdopv :0 mom

32.
8cm
*8

 

 

2383. 83

L $8"
53.4% gem a mo 23802 oﬁ wmoE<

as 95:. 55 8V £9530 cacao sweat. 3:325:80 a av
. €338.53 5X3 £033 3 888m 9.1 mm ~88ng Ala chum—rm

: Emma Em

 

 

44

Section 7. Theoretical Framework of A doption of Hybrid Maize SC513 and Chemical

Fertilizers NPK and Urea

The literature review has shown that adoption studies can be conducted using
frameworks that focus on measurement of variables that inﬂuence adoption, the
innovation-decision process which describes mental processes, and overt behavior, or
combine measurement with components of the innovation-decision process.

Because the present study determines the factors associated with adoption,
including the inﬂuence of the “awareness of improved maize varieties and chemical
fertilizers” and “attitudes toward improved maize varieties and chemical fertilizers”
factors on adoption, the study uses a framework that combines measurement with some
components of the innovation-decision process. The ﬁ'arnework is similar to that used by
Neupane et a1. (2002) and is presented in (Figure 6). Each variable presented in the
conceptual framework (Figure 6) is described below.

Adoption of Hybrid SC513, Chemical Fertilizers NPK and Urea

Adoption is the dependent variable - i.e., the variable explained or predicted by
the independent variables “agro-ecological zone”, “farmer’s age”, “formal education”,
“family size”, “sources of information”, “how-to knowledge”, and “farmers’ attitudes”.
In this study, adoption is deﬁned as the cultivation of SC513 and application of fertilizers
NPK and urea on maize in a given period of time. The period of time in adoption studies
is not predeﬁned. Some adoption studies use a period of time of one year. In
Mozambique, the rates of adoption are very low, which means that the likelihood of
randomly selecting non-adopters would be high. Therefore, a conservative approach was

adopted in this study to set the time of adoption. In this investigation, for prediction

45

proposes, the time of adoption was taken as the period from 1995, when the technological
package Hybrid SC513, NPK and urea, was formally disseminated in Mozambique, to
the year 2005. Adoption was measured by asking farmers the following questions:
“Have you ever cultivated hybrid maize SC513?”, “Have you ever applied NPK and /or
urea on maize?”. Farmers who answered yes were subsequently asked to estimate the

year in which they started using Hybrid maize SC513, and applying NPK and /or urea.

46

 

 

V

 

 

 

08: mo
“cocoa :on a E
388 no .85 98

v52 meouzmtom
Mo gang—mam

ES ﬂmom
.«o gags—=0

 

 

«8: v5 viz
mandate“ 32:85

.38 ceaﬁo
8383‘

   
 

.mp5 Ea v32 .2 wow 382 2.5»: we sauna?» e8 xeoaoESm 3:53:00 .0 Saw?“

 

85 98
VEZ no 380 @338 owBE< o
no.5 338 :o «2: can Mm Z
we 38b“. @339 0253?. o
.m _ ﬂow
8on 088.: c833 3332 o

m _ now no 33380830
cowoscoa 6338 ovBE< o

 

 

 

 

we: can v52

8323.5.“ 3282?
.m _ mom can;

@339 335?.

 

A

 

A

 

 

 

   
   

 

 

8:. Ea viz
€323.59 32823
as m _ new

but? as:

so saunas—8E

mo moon—om

 

 

No.5.— Uﬁd Mamz
agate usage

.2QO can: co
mmoc083<

 

 
 

 

 

owvﬂaoﬁ 8.262 .onm DEB.» £09335 .ow<

 

 

 

 

 

 

£565; 28 muSEEBEV :omwoe Eommo—oooéuwxx

A

   
 

 

 

 

 

47

Agro-ecological Zone

The conditions of the natural environment affect adoption of agricultural
technologies (Rotter and van Keulen, 1997). Adoption patterns may differ signiﬁcantly
between two agro-ecological zones (CIMMYT, 1993). Research suggests that high
levels of adoption of improved maize varieties are more likely to be found among farmers
located in regions with high rainfall. For example, Kaliba et al. (2000) mentioned that
improved maize seeds do better than local varieties in high rainfall regions. Hintze et a1.
(2003) in their study of variety characteristics and maize adoption in Honduras found a
low level of adoption of improved maize among farmers located in a drought-prone
region. From that it can be expected that farmers in Machipanda are more likely to have
cultivated SC513 than farmers in Vanduzi.

Regarding adoption of chemical fertilizer, farmers located in the highlands of
Machipanda are more likely to have applied NPK and urea on maize than farmers in the
lowlands of Vanduzi because the highlands are prone to erosion and loss of soil fertility
due to high levels of precipitation.

Farmer ’s Age

 

Age is another demographic analyzed in adoption studies. In rural Mozambique, the
average age is about 44 years (Trabalho the Inquérito Agricola [TIA] (Agricultural
Survey), 2005). According to CIMMYT (1993), the younger the farmer, the more likely
he/she is to adopt a new technology because he/she has had more schooling (CIMMYT,
1993) than older farmer and is more susceptible to attitude change (Visser and Krosnick,

1998).

48

Hence, the more advanced the farmer’s age, the more likely there is a negative
association with the adoption of improved maize production technology. Some empirical
studies had found a signiﬁcant negative association between farmer’s age and adoption of
improved maize production practices. Abebaw and Belay (2001) surveyed 94 Ethiopian
farmers and found a signiﬁcant negative relationship between age and adoption of high-
yielding maize varieties. Lawal, Saka, Oyegbami, and Akintayo (2004) surveyed 64
farmers in southwestern Nigeria and also found a negative relationship between age and
adoption of improved maize varieties.

Age has also been approached in terms of its importance for agricultural
development as a whole. Scholars in agricultural extension have since been advising
development of extension program for targeting the youth, on the basis of the following
argument:

“. .. young people are the future, and if they learn more about
improved agricultural technology and gain increased conﬁdence and
pride, they can become a major force for long-term agricultural
development, including increased agricultural production and
improved living conditions in the rural areas” (Swanson, Roling, and
Jiggins, 1984, pp. 95).

Thus if agricultural development is the major concern, emphasis should not only be
on the effect of age on adoption of improved technology but also on providing speciﬁc
agricultural development programs targeting the rural youth.

Formal Education

Several empirical studies found that formal education is positively associated with

the probability of adopting agricultural technologies (F eder et al., 1985; Huffman, 1974).

According to Huffman, formal education and extension increase the “allocative ability,”

deﬁned as the human ability to perceive changes in economic conditions (such as relative

49

prices) and to respond efﬁciently. This involves perceiving that change has occurred,
collecting and analyzing useful information, drawing a conclusion, and acting quickly
and decisively (Huffman, 1974). Hence, formal education and extension services
predispose farmers to take an interest in a new technology and make informed decisions
about the technology.

Education would appear to be of particular relevance to the adoption of complex
technologies. For example, CIMMYT (1993) argues that education is expected to be
more signiﬁcant to the adoption of complex technologies such as agrochemicals, than to
improved maize varieties. This view is supported by some empirical studies. Abebaw
and Belay (2001) and Tesfaye and Alemu (2001) found no association between formal
education and adoption of high-yielding maize varieties in Ethiopia. The results are
mixed for chemical fertilizers. A study by Tesfaye, Bedassa, and Shiferaw (2001) found
positive associations between formal education and adoption of chemical fertilizers.
However, Tesfaye and Alemu (2001) found no association between formal education and
adoption of chemical fertilizers.

To some extent, information and extension about technology substitute for formal
education in the rural areas. This may explain the lack of relationships between formal
education and the adoption of improved maize varieties and chemical fertilizers (Abebaw
and Belay, 2001; Tesfaye, Bedassa, and Shiferaw, 2001).

F amily Size
Generally, adoption of new varieties requires increased labor inputs (Feder et al.,
1985). It is assumed that large families provide the labor required for improved maize

production practices. Therefore, family size has been hypothesized to be positively

50

associated with the probability of adoption (Abebaw and Belay, 2001). In rural
Mozambique, average family size is about six members per household (TIA, 2005).
Some empirical studies found a positive association between family size and the
adoption of improved maize varieties (Tesfaye and Alemu, 2001; Tesfaye, Bedassa, and
Shiferaw, 2001). However, there is also evidence of a negative association between
family size and adoption of chemical fertilizers (Tesfaye and Alemu, 2001). According
to Tesfaye and Alemu, lack of funds to purchase fertilizer is responsible for the negative
association between family size and adoption of chemical fertilizer use by smaller
households and non-adoption of fertilizer use by larger households.
Sources of Information

Information received about the technology inﬂuences attitudes (Lichtenberg and
Zimmerman, 1999). Several sources, including extension services, farmers, and
agricultural sales people, provide information on improved maize production practices
(CIMMYT, 1993; Ryan and Gross, 1943), raising farmers’ awareness.

Although farmers and inputs dealers may provide information about the
technology, it is the extension service’s responsibility to deliver information inputs to
farmers and eventually persuade farmers to adopt the technology being disseminated.
Information can be of many kinds, including estimates of future prices for farm products,
new research products, and knowledge about the timing and intensity of fertilizer use
(Anderson and Feder, 2004).

Research found a positive association between extension and adoption of
technology. Pannell et al. (2006) mentioned that farmers are more willing to adopt new

technology if they respect and trust the source, such as extension agents. Some empirical

51

studies on determinants of adoption of improved maize varieties and chemical fertilizers
provide mixed evidence. Abebaw and Belay (2001), Tesfaye, Bedassa, and Shiferaw
(2001), and Tesfaye and Alemu (2001), on the basis of the number of contacts with
extension, found a positive association between extension and adoption of improved
maize. In the same study, Tesfaye and Alemu (2001) also found a negative association
between extension and the adoption of chemical fertilizers.

How-to Knowledge

How-to knowledge is motivated by awareness knowledge and consists of
information necessary to use the technology properly (Rogers, 1995). As with
information about the technology, how-to knowledge inﬂuences behavior through
attitude formation (Lichtenberg and Zimmerman, 1999). An inadequate level of how-to
knowledge is likely to lead to rejection and discontinuation of technology (Rogers, 1995).
Farmers ’ Attitudes

Farmers’ attitudes toward agricultural technology and their inﬂuence on adoption are
not yet widely studied. For this study, the following attitudinal dimensions were created:
attitude toward production characteristics of SC513; attitude toward income from SC513;
attitude toward effects of NPK and urea on maize crop; and attitude toward costs of NPK
and urea.

Existing literature explains that adopters tend to hold positive attitudes toward the
technology (Rogers, 1995). Empirical logit estimates have indicated that farmers’
positive attitudes toward technology favor the adoption of it (Chilonda and Van
Huylenbroeck, 2001). While studying perceptions and attitudes toward soil fertility

technologies in western Africa, Enyong et al. (1999) found that farmers perceive urea as

52

being more costly than NPK because it had less residual effect and must be purchased
yearly. Hence, for the present study, farmers holding positive attitudes toward the
characteristics of and income from improved maize variety SC513, and attitudes toward
effects of, and costs of NPK and urea are expected to adopt them. And between NPK and
urea, farmers are more likely to have positive attitudes toward chemical fertilizer NPK
because it has more residual effect and farmers may not be required to purchase NPK

yearly.

Section 8. Study Hypotheses

From the descriptions of variables that inﬂuence adoption of improved maize

varieties and chemical fertilizers, the following study hypotheses were formulated:

H3: Respondents in the highlands of Machipanda are more likely to adopt
improved maize variety SC513.

H4: Respondents in the highlands of Machipanda are more likely to adopt NPK
and urea.

H5: Farmer’s age is negatively associated with adoption of improved maize
SC513 and chemical fertilizers NPK and urea.

H6: A farmer’s household size is positively associated with adoption of SC513
and NPK and urea.

H7: A farmer’s level of education is positively associated with adoption of SC513
and NPK and urea.

H8: Extension is positively associated with adoption of SC513 and NPK and urea.

H9: How-to knowledge is positively associated with adoption of SC513 and NPK

and urea.

53

H10: Positive attitudes toward SC513, NPK and urea are positively associated

with their adoption.

Section 9. Logistic Model

The measurement component of the analytical framework is described by the
logistic model. The logistic model is a very popular statistical technique used for
studying factors associated with adoption of technology (N eupane et al., 2002; Chilonda
and Van Huylenbroeck, 2001; CIMMYT, 1993). According to Hosmer and Lemeshow
(2000), in many ﬁelds, the logistic model is the standard method of analysis when the
outcome variable is dichotomous. Logistic regression differs from multiple regression in
being speciﬁcally designed to predict the probability of an event occurring (i.e., the
probability of an observation being in the group coded 1) (Hair et al., 2005). In the ﬁeld
of agriculture, adoption of technologies is often measured quantitatively as a
dichotomous response variable (0 = non-adoption of innovation, and 1 = adoption of
innovation).

The logistic regression model characterizing adoption of SC513 or NPK or urea

by the sample of households is speciﬁed as follows:

e or + BIX1+ Bzxz +~-+ BPXP
1t = (eq. 1.0)
1+ e 0‘ + BIXI + B2X2+...+ BPXP

 

Where:

54

it is the actual proportion of farmers adopting the technology for particular values
of independent variables X], X2, ..., Xp, that inﬂuence adoption of SC513 or NPK or

urea.

[51, [32 ,. . ., Bp, denote the regression coefﬁcients associated with independent

variables X1, X2, ..., Xp, that inﬂuence adoption of improved seed SC513 or chemical

fertilizers, NPK or urea.

From equation 1.0, it is possible to arrive at a simple linear regression equation
through logit transformation (eq. 2.0). Instead of working directly with it, we work with
a transformed value of it (Chatterjee, Hadi, and Price, 2000; Hosmer and Lemeshow,

2000). After logit transformation, the logistic model takes the following linear form:

log [113/ (1 - 1r)] = or + [31X1+ B2X2 +...+[3po (eq. 2.0)

Because farmers may decide to adopt the major technical innovation from the
package (e. g., improved maize seeds) and choose to adopt other complementary
components (e.g., fertilizer or pesticides) as they learn by doing (Kaliba et al., 2000), the
logistic model in equation 2.0 is applied separately for each technology - hybrid maize
SC513, NPK, and urea.

Interpretation of the Regression Coefficients
The transformed equation 2.0 shows that the estimated coefﬁcients do not directly
indicate the effect of change in the corresponding explanatory variables on probability

(1:). Rather, the coefﬁcients represent the change in the logarithm of the odds of success

55

on the dependent variable for each increase of one unit on the independent variable. A
positive coefﬁcient means that the logarithm of the odds of success increases as the
corresponding independent variable increases (N eupane et al., 2002).

Logistic regression coefﬁcients are difﬁcult to interpret in their original form
because they are expressed in terms of logarithms when we use the logit as the dependent
variable. Therefore, we use the exponentiated logistic coefﬁcient (Exp (3)), which gives
us the ratio of the probabilities (the odds) (Hair et al., 2005). For example, if Exp ([3) = 2
(i.e., odds ratio equal 2), this means that each one unit increase in the independent
variable associated with coefﬁcient [3 results in doubling the odds of success in the
dependent variable.

For the interaction term, the coefﬁcient [3 indicates the difference between the rate
of change for the group associated with B and the rate of change for the reference or ﬁnal
group. If the interaction term is statistically signiﬁcant, it indicates that the difference
between the two groups changes in function of the values of the second variable. And if
the sign of the interaction term is positive, it means that the difference between the two

groups increases as the value of second variable increases.

56

CHAPTER III

METHODOLOGY

This chapter is organized in ﬁve sections. Section 1 describes the study location.
Section 2 presents the research design. Section 3 presents the steps followed for
examining data. Section 4 describes the statistical techniques used to analyze data.

Section 5 discusses the study limitations.

Section 1. Study Area

The study was conducted in two of Manica District’s ﬁve administrative posts:
Machipanda and Vanduzi (Figure 7). Machipanda and Vanduzi were deliberately
selected for several reasons. First, Machipanda and Vanduzi are located in one of the
major agricultural provinces in Mozambique with maize-based farming systems. Second,
the maize growers in Machipanda and Vanduzi offer an opportunity to learn about
adoption of improved maize varieties and chemical fertilizers because the two locations
were target areas of the DNER/SG2000 program for dissemination of SC513, NPK, and
urea. Third, Machipanda and Vanduzi are located in different agro-ecological regions.

This fact allows comparisons of highland and lowland growing areas.

57

PROVINCIA DE MANICA
Distrito de Manica

 

  
  

 

 

   

 

  

 

 

\ K I
.2
l
1..
\
Easel. 1:550 .000 K
4 0 12 10 m Klomclu'»
c
(.
- i
k.
H
l
\
nMBAawm (
.u,"_ﬁ
IDOLA
.I'I- '
- ‘1'
l'
«*4
‘.
2
S
\ ..r‘ 9... w'
~./ \1 V’gu
SUNSSUNDENOA
.5
.

 

 

Source: National Directorate of Mapping Services (DINAGECA) 2006.

Figure 7. Map of the Manica District Showing the Five Administrative Posts (white dots),
including the Two Research Sites, MACHIPANDA and VANDUZI Administrative

Posts.

58

Rainfall Pattern for the Entire District

The pattern of rainfall for the entire district, shown in Figure 8, was used as proxy
for the rainfall pattern in Vanduzi and Machipanda. The rainfall pattern in the district is
reliable. Nevertheless, drought spells can occur for short periods (Bias and Donovan,
2003; ICRISAT, 2004). When this happens, parts of Machipanda and Vanduzi are
affected. To cope with the drought, farmers plant early-maturing and drought-tolerant

maize varieties (ICRISAT, 2004).

Maize growing season precipitation in mm

 

5' 1951-1980

I l995-2006

 

 

 

  

DEC JAN FEV MAR APR

Source: National Weather Forecast Services, 2006.

Figure 8. The Unimodal Pattern of Rainfall in the Manica District.

Administrative Post of Machipanda

Machipanda is located in the area bordering the Republic of Zimbabwe, in agro -
ecological region 10. This is the high altitude region of the Manica district, with altitude
greater than 1,000 meters above sea level. The annual rainfall is greater than 1,200 mm,
and the average temperature varies between 15 and 22. 5 degrees Celsius. The soil types

are mainly ferrasols of heavy texture. Overall moderate to high nutrient mining may

59

occur because of low levels of inputs used for crop production (Uaiene, 2004). Also soil
erosion and loss of soil fertility due to the high levels of rainfall (INIA, 1997) are major
constraints to agricultural cultivation. Soil erosion control and soil fertility management
are important for maintaining the productivity of the soils.

Land use in Machipanda is characterized by agricultural cultivation and forest
plantation. Crop cultivation is the main activity for individual households, and maize is
the dominant crop with a high production potential. There is mainly one maize crop per
year. Maize is sown in the rainy season, from October to December, and harvested
between April and May. With rain in May, a second maize crop can be planted in the dry
season or with irrigation in selected areas (Bias and Donovan, 2003).

The maize farming system is mainly rain-fed, and maize production is
characterized by manual cultivation and animal traction. Farmers cultivate several plots
(between 0.5 and 3 hectares) with local maize (Chimanhica, Candgere) and improved
maize varieties (mostly Hybrid maize SC513, SC501, PNR30G97) intercropped with
pumpkin, beans, and cowpea. Hybrid maize seed is obtained mainly from informal
traders. Other acquisition methods include buying improved maize seed from a local
seed depot or street market, saving it from one year to the next, or btaining it from other
farmers/relatives or extension services. Some farmers use fertilizer for maize production.
Farmers obtain fertilizer mainly from informal traders.

In Machipanda, maize produced (output) is mainly used for home consumption.

Farmers have access to two open markets. One open market is located nearby the
center of the administrative post of Machipanda; the second and major open market is

located in the center of the administrative post of Manica (about 15 minutes’ drive from

60

Machipanda). Some farmers participate in these markets with maize grain, vegetables,
and tubers. Overall the terrain and the elevation make it difﬁcult to transport products to
local markets.

Administrative Post of Vanduzi

 

Vanduzi is located in the lowlands, in the R4 agro-ecological region, the most
common agro-ecological region in the Manica district. It has an altitude between 200 and
1,000 meters above sea level. The annual rainfall is between 1,000 and 1,200 mm.
During the growing season the average temperature varies between 17. 5 and 22 degrees
Celsius. The soils are generally light, with some occurrence of heavy soils. Overall
moderate to high nutrient mining may occur because of low levels of inputs used for crop
production (Uaiene, 2004).

'As in Machipanda, crop cultivation is the main activity, and maize is the
dominant crop with a high production potential. There is mainly one maize crop per
year. Maize is sown in the rainy season, from October to December, and harvested
between April and May. With rain in May, a second maize crop can be planted in the dry
season or with irrigation in selected areas (Bias and Donovan, 2003). The maize farming
system is mainly rain—fed, and maize production is characterized by manual cultivation.
Farmers cultivate several plots (between 0.5 and 3 hectares) with local varieties
(Chimanhica, Candgere) and improved maize varieties (mostly the open-pollinated
varieties Matuba, Sussuma, and PAN 67) intercropped with pumpkin, millet, and
sorghum. Improved maize seed is obtained mainly from the local seed depot. Other
acquisition methods include buying improved maize seed from a street market, saving it

from one year to the next, and obtaining seed from other farmers/relatives or extension

61

services. Some farmers use fertilizer for maize production. Farmers obtain fertilizer
mainly from informal traders and tobacco companies, with which farmers have contract
growing schemes.

In Vanduzi, maize produce is important for both home consumption and as a
source of income. Farmers have access to one open market located in the center of the
administrative post. Here, some farmers sell maize grain, vegetables, tubers, and
charcoal. Another important market for maize grain is provided by the largest
wholesaler/retailer in the administrative post. Farmers bring their own small sacks with
maize grain and sell directly to the wholesaler/retailer. This exchange was not commonly
observed in Machipanda.

Machipanda and Vanduzi Compared
Table 7 summarizes the major agroclimatic and farming characteristics of

Machipanda and Vanduzi, and compares the two administrative posts.

62

Table 7. Key Agro-climatic and Farming Characteristics of Machipanda and Vanduzi.

 

Key characteristics Machipanda Vanduzi
(Highlands) (Lowlands)

Altitude (m) 900 - 1,500 200 - 1,000

Rainfall (mm) 1,000 - 1,500 1,000 - 1,200

Soil quality Moderate to high Moderate to high nutrient
nutrient mining may mining may occur because of
occur because of low low levels of inputs
levels of inputs
Soil erosion and loss of
fertility

Land use Agriculture Agriculture
Forest plantation

Maize-based farming system Rain-fed Rain-fed
Hybrid maize OPV and Hybrid
Intercropping (maize, Intercropping (maize,
pumpkin, beans) pumpkin, sorghum and
Manual cultivation millet)
Animal traction Manual cultivation

Major use of maize produce

Main source of improved
seed (private sector)
Main source of chemical
fertilizer (private sector)

Market for seed grain

 

Maize mainly for home
consumption

Informal traders

Informal traders

Open market

63

Maize for home consumption
and sale

Seed depot
(wholesalers/retailers)

Informal traders
Tobacco company

Open market
Wholesaler/retailer

Section 2. Research Design

This study involved a cross-sectional survey (Babbie, 1998) with randomly
selected households growing maize in the Machipanda and Vanduzi administrative posts
of the Manica district. Rural households producing maize in the highlands of
Machipanda (N = 3,133) and the lowlands of Vanduzi (N = 5,175) represented the
population of interest in this investigation (INE, 1997). The sampling consisted of 120
farmers in Machipanda and 173 farmers in Vanduzi. Data were collected between April

and May 2006 using personal interviews.

Sampling Design

A multistage sampling design (Kendal and Buckland, 1971) was used with the
help of the Mozambique National Institute (INE). The sampling stages are presented in
Table 8. The Primary Sampling Unit (PSU) was deﬁned as the administrative post. At
this stage, Machipanda and Vanduzi were purposively selected. The Secondary
Sampling Unit (SSU) was the locality, which is a rural subdivision of the administrative
post. All localities in Machipanda and Vanduzi were included in the sample. The village
constituted the Third Sampling Unit (TSU). A total of 26 villages (10 villages in
Machipanda and 16 villages in Vanduzi) were randomly selected. The household was
deﬁned as the Ultimate Sampling Unit (USU). Within each village, 8 t012 households
(hh) were randomly selected, using the random walk method (Nichols, 1991; Kendal and

Buckland, 1971).

64

 

 

 

8m 8% 3. m N _ 3,;
a MS 850 §9§§> 55> was: 8
E m: osasom 8 mm 082N35> E23, 852 mm
_ _ 5 $220 QBE§§> §2a> 83% E
S 5. 3&5 ouom-§w§> m§e§> 8%: a
S a: madam 08...-_N%§> Baas, 8%: mm
_ _ $2 acauﬁqgenu 08e§u§> 53> 8%: a
2 mg «was? 8»...-§§> 565$ 852 em
_ _ mm owﬁoaaaﬁao 039§u8> n35.» 8:32 2
2 R oaessanz amazes; §u§> «can: 2
m S 85%: Engage m§e§> “0%: t
2 RN 880 am ”amaze Engage _§§> «232 E
2 8 BE: Engage 53> nova: 2
2 SN sausage 225mg “35> 852 z
2 :N 3338532 2235 53> «can: 2
4 am 05682 censum 08898 53> News: 2
: NS E5. 8 h eouowno m§a§> 852 :
N_ mmm omega; owaost avgmﬁowE 8:32 3
S mom mocwoﬁﬁmoog £802 .< owcowaz muﬁaﬁomz 8:52 mo
S or 35:. 8 mm 88%: €538: 852 we
2 an. 88:09: 35: «Bang: «can: 8
2 ”mm mango“? Suﬁ: €328: mama: 8
2 m3 255:2 3E2 «cannon: song: 8
2 E 252 352 magmas: «can: 3
S MEN evanescence 352 3:828: SEE 8
2 ms. «26 352 $8282 8:32 No
Q 8 888:: 552 88?an 83.2 s
BEEea mm 8388 m: 83.5 3%.: mom .§< aroma e 82

 

 

.mnmom o>u§££§< 3653 28 3:89:82 5 mEonomsom mo mamas—am owSmEEz .m 3an

65

Development of the Survey Instrument

Several materials were used to assist in development of the survey instrument.
The materials included notes on maize production (Hoeﬁ et al., 2000), articles on
farmers’ preferences for maize varieties (Hintze et al., 2003), questionnaires on farmers’
attitudes and adoption of improved agricultural technology (Rogers and Havens, 1961;
Diallo, 1983; Jeje at al., 1998; INE, 2001), and scales for the measurement of attitudes
toward farming and attitudes toward any practice (Shaw and Wright, 1967).

The survey instrument contained 26 items distributed in ﬁve sections, as

presented in Table 9.

Table 9. Structure of the Survey Instrument.

 

Section Information collected Number of
items

1 Knowledge and adoption of improved maize varieties
Farmers’ attitudes toward improved maize varieties 1

3 Knowledge and adoption of chemical fertilizers
(N PK/urea)

4 Farmers’ attitudes toward chemical fertilizers (NPK/urea) 1

5 General information and demographics

Total 26

Section 1 included items about knowledge and adoption of improved maize
varieties. The items addressed farmers’ experience (years) in cultivating maize,
awareness of improved maize varieties, type of maize varieties cultivated in the last
growing season between October and December 2005, number of plots under maize
cultivation, proportion of land cultivated to improved varieties, the tillage and cropping
system used, adoption and discontinuation of improved maize varieties, and type of credit

for improved seed.

66

Section 2 included items about farmers’ attitudes toward improved maize
varieties. In this section, farmers were asked their degree of agreement with sixteen
attitudinal statements on Hybrid maize variety SC513. Thirteen statements expressed
positive attitudes toward SC513, and three statements expressed negative attitudes toward
SC513. Farmers’ degree of agreement with each statement was measured using a S—point
interval scale with 1 = Strongly Agree, 2 = Agree, 3 = Neutral, 4 = Disagree, and 5 =
Strongly Disagree.

Section 3 included items about knowledge and adoption of chemical fertilizers
(N PK/urea). In this section, the information solicited included awareness of NPK and
urea, type of fertilizer applied on their maize crop in the growing season October-
December 2005, proportion of land fertilized, adoption and discontinuation of NPK and
urea, and type of credit for use of chemical fertilizers.

Section 4 included items about farmers’ attitudes toward NPK and urea. In this
section, farmers were asked their level of agreement with eight attitudinal statements on
NPK and eight attitudinal statements on urea. Among the eight statements, four
statements expressed positive attitudes toward NPK and urea, and four statements
expressed negative attitudes toward NPK and urea. Farmers’ degree of agreement with
each statement was measured using a 5-point interval scale with I = Strongly Agree, 2 =
Agree, 3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree.

Section 5 included general information and demographics questions. Information
solicited included the demographic variables of the farmers and their immediate families,
assistance from extension services in the past 12 months, and an Open question on maize

production.

67

Validity of the Survey Instrument

Content validity (Litwin, 2003) was ascertained through review of the instrument
by the dissertation supervisor, guidance committee members and an expert on agricultural
survey in the Ministry of Agriculture in Mozambique. These experts provided feedback
on the survey schedule. On the basis of their review, modiﬁcations were made. These
included additional attitudinal statements and improved ﬂow of the questions through the
use of conditional branching for statements that did not apply for the respondent (Alreck

and Settle, 2004).

Reliability of the Survey Instrument

Cronbach’s alpha was performed to measure the reliability of the statements used
to assess farmers’ attitudes toward improved maize variety SC513 and NPK and urea. A
Cronbach’s alpha value of 0.634 was estimated for the reliability of 3 statements on
attitudes toward marketability of SC513, and a Cronbach’s alpha value of 0.615 was
estimated for the reliability of 3 statements on attitudes towards production
characteristics of SC513. Cronbach’s alpha values of 0.536 and 0.677 were estimated for
four statements on NPK and four statements on urea. Most authors of measurement
textbooks do not set a cut-off value for acceptable and unacceptable reliabilities. In some
cases, reliability value of 0.5 or 0.6 is acceptable; in others, a value of 0.9 is barely
acceptable (Kerlinger and Howard, 2000). Cosidering the exploratory nature of this

study Cronbach’s alpha values of 0.677 and 0.536 were considered adequate.

68

Selection of Interviewer

One interviewer (born and raised in Manica) was contracted on March 17, 2006,
two weeks before data collection. The interviewer held a bachelor’s degree in agronomy
from Eduardo Mondlane University. She knew the research protocols and was willing to
follow the instructions and adhere to the directions. The assistant also possessed

independent views of the technology under study.

Informing Local Authorities

The researcher introduced herself and the interviewer to the local authorities. The
district director of agriculture, the presidents of the administrative posts, and the villages’
representatives were informed about the research objectives and the survey schedule

(dates).

Interviewer Training

Before data collection, the interviewer received three days of training on survey
interviewing procedures. Five aspects were stressed during the training. First, it stressed
the academic objective of the study, the letter of consent, and voluntary participation of
respondents. Second, the training stressed the screening question, i.e., the importance of
obtaining responses from a single respondent in each household. The interviewer was
instructed to collect data from one respondent who was the head of the household (or
wife of the head of the household) and was involved in maize cultivation. In cases where
a couple was approached, the couple decided who should provide the answers. Third, the
training stressed taking time to pause, listen, perceive the respondents’ answers and

comments, and reread the question if necessary. Fourth, the training stressed the use of

69

the 5-point scale to measure attitudes toward SC513, NPK and urea. Fiﬁh, the training
stressed consistency of the sequence of questions. The interviewer had to follow the
same sequence every time when asking questions. The training also included a role-play
practice technique, whereby the interviewer interviewed the researcher using both

Portuguese and local languages.

Oral Translations

Some respondents could not speak Portuguese ﬂuently; therefore, the questions
needed to be translated into Chimanhica, the local language. Although the researcher
understood Chimanhica, her ability to speak it was not sufﬁcient to translate directly from
Portuguese to Chimanhica during the interviews. Therefore, to ensure accurate
translations, three local people ﬂuent in both the Portuguese and Chimanhica languages
were contracted and thoroughly trained in translating the questions and the 5-point scale
for the attitudinal statements. During the interview, the researcher read each question in

Portuguese, which was translated into the Chimanhica language for the respondent.

Pilot Testing

The survey was given to twenty-two farmers from the Machipanda and Vanduzi
administrative posts. Thirteen respondents were male and nine were female. In the last
rainy season, twenty respondents planted hybrid and/or OPV maize, and two respondents
applied NPK (12-24-12) and urea. The respondents demonstrated a willingness to
participate and felt free and comfortable to answer questions regarding production of
their maize crop. The respondents considered the vocabulary and items very simple,

appropriate, and sensitive to their culture.

70

Data Collection and Supervision

Data were collected through face-to-face interviews. The interviews were
administered by the researcher and the ﬁeld assistant separately. The interviews were
conducted from 7 am. to 2 pm. and from 2:30 pm. to 7 pm. About twelve interviews
were completed per day. Each interview took about ﬁfty ﬁve minutes to complete. In
each household data were collected from one respondent, who was the head of household
(or wife of the head of household) and involved in maize cultivation. At the end of the
day, the interviewer reviewed the paths followed for the selection of households to ensure
that the random walk procedure was followed correctly. The researcher also counted and
checked the completed questionnaires to make sure that data were recorded correctly and

the questionnaires were signed.

Section 3. Data Examination

Data examination is the initial step in any analysis. In this study, data
examination was performed using the Statistical Package for Social Sciences (SPS S)
version 15 (SPSS Inc., 2006) and AMOS Graphics 7. The following aspects were
addressed in data examination:

Missing Data

First, the missing data were examined. About 6.4 percent of the data on attitudes
toward improved maize variety SC513 were missing. These data were not included in
the subsequent analysis. On attitudes toward NPK and urea, there were 15 percent (or 44
out of 293 respondents) and 13.7 percent (or 40 out of 293 respondents) missing data,
respectively. For these data, replacement values were estimated through maximum

likelihood procedure. This procedure attempts to model the processes underlying the

71

missing data and to make the most accurate and reasonable estimates possible (Hair et al.,
2005). Replacement values were estimated for only 31 respondents who at least had
heard about the fertilizers NPK and urea.
Multivariate Normality Assumptions
Second, data were tested to ckeck if multivariate normality assumptions were
reasonably met in order to run conﬁrmatory factor analysis (CFA). Frequencies and
bivariate scatter plots were examined to check for normality (and homoscedascity)
assumptions. The data indicated no gross departures from normality. The inspection of
bivariate scatter plots indicated no curvelinearity, meaning that data approximated
linearity. For most of the variables, the assumption of constant variance was not grossly
violated.
Summated Scales
On the basis of the literature review (Chapter 2), the following was hypothesized
about the attitudinal dimensions (factors):
H1: Attitudes toward production characteristics of SC513 are positively associated
with attitudes toward income from SC513.
H2: Attitudes toward effect of NPK and urea on maize yield are negatively
associated with attitudes toward costs of NPK and urea.
Thus, the third step in data examination involved performing exploratory and
conﬁrmatory factor analysis to test these hypotheses and create summated scales of
attitudes toward SC513, NPK, and urea. The responses to attitudinal items were subject

to exploratory factor analysis to uncover the factors in attitudes. The factors were

72

supported using a conﬁrmatory factor analysis (Grimbeek and Nisbet, 2006; Borjesson,
Aarons, and Dunn, 2003; Hemmerlgarn et al., 1995).

The results of exploratory factor analysis are presented in Tables 10-12. For the
scale on attitudes toward maize Hybrid variety SC513 (Table 10), the two-factor model
was satisfactory. The solution had Chi-square = 3.845 and P = 0.427 (P>.05), 2
Eigenvalues greater than 1, and 6 percent residuals correlations with absolute value

greater than .05.

Table 10. Factor Loading and Coefﬁcients of Reliability on Attitudes Toward SC513.

 

 

Oblique Rotated
Loadings
Factor

1 2
Grain from SC 513 is easy to sell. .758
Fresh maize (roasting cobs) of SC 513 is easy to sell. .638
Cultivation of improved maize variety SC 513 is a waste 408
of time and money. '
When rain is scarce, local maize variety Chimanhica has 53 5
better production than SC 513. '
Grain from SC 513 is good for milling. .498
Seed from SC 513 has good germination. .545

Cronbach ’s al ha 634 .615

On the basis of the high loadings (.758 and .638) for the statements “Grain from
SC 513 is easy to sell” and “Fresh maize (roasting cobs) of SC 513 is easy to sell”, factor
1 was named Attitudes toward marketability of improved maize hybrid variety SC513.
Factor 2 was named Attitudes toward production characteristics of improved maize
hybrid variety SC513, because high loadings (.545 and .535) were observed for the
statements “Seed from SC 513 has good germination” and “When rain is scarce, local

maize variety Chimanhica has better production than SC 513”.

73

As presented in Table 10, each of the two factors on SC513 had three items.
Hence, for each of the two factors (Factor 1 and Factor 2) on improved maize variety
SC513, the summed score across items for a given respondent ranged from 3 to 15, with
scores lower than 9 indicating favorable attitudes and scores greater than 9 indicating
unfavorable attitudes toward selling SC513.
For the scale on attitudes toward NPK, one factor model was satisfactory (Table 11).
This solution had 1 Eigenvalue greater than 1, and 16 percent residuals with absolute

value greater than .05.

Table 11. Factor Loadings and Coefﬁcient of Reliability on Attitudes Toward NPK.

 

Attitude statements (Items) Factor loadings
NPK is good for maize. .851
NPK increases maize yield. .703
NPK is not good for the soil. .164
NPK is a waste of time and money. .425
Cronbach ’3 alpha .536

As Table 5 illustrates, the attitudes toward the effects of NPK on maize production
were the most dominant. The statements “NPK is good for maize” and “NPK increases
maize yield” had the highest loadings (.851 and .703).

The factor Attitudes Toward NPK had four items, so the summed score across items
for a given respondent ranged from 4 to 20, with scores lower than 12 indicating
favorable attitudes and scores greater than 12 indicating unfavorable attitudes toward
NPK.

For the scale on attitudes toward urea, one factor model was found reasonably
satisfactory (Table 12). This solution had 1 Eigenvalue greater than 1 and reasonable

sizes of factor loadings.

74

Table 12. Factor Loadings and Coefﬁcient of Reliability on Attitudes Toward urea.

 

Attitude statements (Items) Factor loadings
Urea is good for maize. .961
Urea increases maize yield. .878
Urea is not good (harmful) for the soil. .292
Urea is a waste of time and money. .382

Cronbach ’s al ha .677

As Table 12 illustrates, the attitudes toward the effects of urea on maize
production were the most dominant. The statements “Urea is good for maize” and “Urea
increases maize yield” had the highest loadings (.961 and .878).

The factor Attitudes Toward urea had four items, so the summed score across
items for a given respondent ranged from 4 to 20, with scores lower than 12 indicating
favorable attitudes and scores greater than 12 indicating unfavorable attitudes toward
urea.

The results of conﬁrmatory factor analysis obtained from AMOS Graphics 7 are

presented in Table 13.

Table 13. Fit Statistics for Conﬁrmatory Analytic Models.

 

Model CMIN (X 2) df P CFI RMSEA
Two-factor model for improved 7.550 8 .459 1.00 .000
maize variety SC513

One-factor model for NPK 7.285 2 .026 .968 .101
One-factor model for urea 31.267 2 .000 .929 .237

 

The two-factor model for improved maize variety SC513 has an excellent ﬁt, as
indicated by a non-signiﬁcant X 2 = 7.550, p >005. Comparative index (CFI) is greater
than 0.90, and the root mean square error of approximation (RMSEA) is lower than 0.05.

The one-factor model for NPK has a good ﬁt, as indicated by a non-signiﬁcant X 2 =

75

7.285, p >0.01. The comparative index (CF I) is greater than 0.90. Although the RMSEA
is greater than 0.05, the 90 percent conﬁdence interval for RMSEA is 0.030 to 0.184.
This interval includes 0.05 criterion which indicates “reasonable” errors of approximation
in the population. The one-factor model for urea has reasonable ﬁt. The comparative
index (CFI) is slightly greater than 0.90. This means that although other ﬁt statistics
(CIMIN and RMSEA) suggest a poor ﬁt, the model is still better than the complete or
baseline model.
Variance Inﬂation Factor (VIF)

Fourth, Variance Inﬂation Factor (VIF) estimates were examined for collinearity
diagnostic. Most of VIF estimates had values less than 2, which indicate no serious
problems of collineraity (Appendix F). Moreover, the regression output did not show

large estimated standard errors (Hosmer and Lemeshow, 2000).

Section 4. Data Analysis

This section describes the analysis of data used to address the speciﬁc research
ojectives. The data were collected using personal interviews. A S-point Likert scale with
I = Strongly Agree, 2 = Agree, 3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree
was used to measure farmers’ attitudes toward improved maize variety SC513 and
chemical fertilizers NPK and urea. Data were analyzed using SPSS 15 (SPSS Inc.,
2006). The steps and statistic techniques used for data analysis are described below.

Descriptives

First, descriptive statistics, percentage, mean, and standard deviation were used to

describe:

76

a) characteristics of respondents (farmers). The variables analyzed were: “Age”,
“Gender”, “Level of education”, “Family size”, “Source of information”, “Knowledge
advantages of improved maize” and “Knowledge of methods to apply fertilizer”.

b) Farmers’ attitudes toward marketability of improved maize Hybrid variety
SC513. The following items were analyzed: “Grain from SC 513 is easy to sell”, “Fresh
maize (roasting cobs) of SC 513 is easy to sell” and “Cultivation of improved maize
variety SC 513 is a waste of time and money”.

c) Farmers’ attitudes toward production characteristics of improved maize Hybrid
variety SC5 1. The following items were analyzed: “When rain is scarce, local maize
variety Chimanhica has better production than SC 513”, “Grain from SC 513 is good for
milling” and “Seed from SC 513 has good germination”.

(1) Farmers’ attitudes toward NPK. Four items were analyzed, namely: “NPK is
good for maize”, “NPK increases maize yield”, “NPK is not good for the soil”, and “NPK
is a waste of time and money”.

6) Farmers’ attitudes toward urea. The following items were analyzed: “Urea is
good for maize”, “Urea increases maize yield”, “Urea is not good for the soil”, and “Urea
is a waste of time and money”.

1) Pattern of adoption of improved maize Hybrid SC513, NPK, and urea. The
variables analyzed were: “Year started planting SC513”, “Year started applying NPK on
maize”, and “Year started applying urea on maize”.

T—test and oneway ANO VA

Second, T-test and oneway AN OVA test were used to compare:

77

a) Mean attitude scores between highlands and lowlands. This analysis involved
the following factors and variables: farmers’ attitudes toward marketability of improved
maize variety SC513, farmers’ attitudes toward production and consumption
characteristics of improved maize variety SC513, farmers’ attitudes toward chemical
fertilizer NPK, farmers’ attitudes toward the chemical fertilizer urea, and agro-ecological
zone.

b) Mean attitude scores among characteristics of respondents within the study
locations. This analysis involved the following factors and variables: farmers’ attitudes
toward marketability of improved maize variety SC513, farmers’ attitudes toward
production and consumption characteristics of improved maize variety SC513, farmers’
attitudes toward chemical fertilizer NPK, farmers’ attitudes toward the chemical fertilizer
urea, age, gender, and, source of information.

Multiple Logistic Regression

 

Third, a logistic regression was performed to determine factors associated with
adoption of the improved maize variety SC513 and chemical fertilizers in the highlands

and lowlands. The variables analyzed are presented in Table 14.

78

Table 14. Description of Variables Entered in the Multiple Logistic Regression.

 

Variable name Variable type Description

Agro-ecological region (XI) Dichotomous 0 = lowlands (Vanduzi), 1 = highlands
(Machipanda)

Age (X2) Continuous Respondent’s age (years)

Family size (X3) Continuous Number of members in the family

Level of education (X4) Dichotomous 0 = illiterate, 1 = some schooling

How-to knowledge (X5) Dichotomous 0 = not knowledgeable, 1 =

knowledgeable
Source of information (X6) A multiple For improved maize SC513:
category 1 = neighbors, 2 = market, 3 = extension
variable For chemical fertilizers:

1= market, 2 = extension, 3 = neighbors
Attitude (X7) Interval Factor scores for attitude toward

marketability of grain of improved maize

SC5 1 3

Factor scores for attitude toward

production traits of improved maize

SC513

Factor scores for attitude toward NPK
Factor scores for attitude toward urea

Section 5. Study Limitations

Using the interview technique for data collection has limitations. One major
limitation is the interviewing error (Alreck and Settle, 2004). For this study, some error
and bias might have occurred because of the use of an interviewer and translators. To
control these errors, the interviewer received training on interviewing procedures, and the
interviewing process was closely monitored. The interpreters were instructed very
thoroughly to give oral translations that were as close as possible to the original

questions, and the researcher monitored the translation process. However, there could be

79

some translator bias in the translation process. Although the researcher
understood Chimanhica, limited familiarity with the Chimanhica language could have
resulted in her missing a few words expressed by the respondents. Interviews with
farmers were conducted by the researcher with the help of a research assistant. Both
interviewed the farmers together while pilot testing the survey instrument. During the
actual data collection, the interviews were conducted independently. Thus, the two
interviewers might have asked questions in a different manner that may have resulted in
some errors.

The attitude scales in this study were used for the ﬁrst time to assess attitudes of
farmers growing maize in Machipanda and Vanduzi in Mozambique. The attitudes
toward improved maize variety SC513 and chemical fertilizers NPK and urea expressed
by respondents in Machipanda and Vanduzi may not generalize to other populations.

As mentioned, for practical reasons, this study considered a ten-year period as the
time of adoption. This period is relatively large (one year is used in some other studies).

Therefore, future studies should try to use relatively short periods such as one year.

80

CHAPTER IV

RESULTS

The results presented in this chapter are based on data gathered from personal
interviews with maize growers in the Machipanda and Vanduzi administrative posts.
Data were analyzed using SPSS 15, and ﬁndings are presented according to each speciﬁc

research objective.

Research Objective 1

The ﬁrst objective of this study was to describe the characteristics of respondents
from the Machipanda and Vanduzi administrative posts. The characteristics of

respondents in the two study locations are presented in Table 15.

Characteristics of Respondents

As shown in Table 15, the majority of farmers in Machipanda (54.2 percent) and
Vanduzi (51.4 percent) were less than 44 years old. Similarly, the majority of farmers in
Machipanda (65.8 percent) and Vanduzi (64.7 percent) had family sizes of about eight
members. Most respondents in Machipanda (54.2 percent) were female, and most

respondents in Vanduzi (61.8 percent) were male.

81

Table 15. Characteristics of Respondents in the Machipanda and Vanduzi Administrative
Posts in the Manica District.

 

 

 

 

Characteristics Study location
Machipanda Vanduzi
Administrative Post Administraive Post
Frequency Percent Frequency Percent
Age (Years) (n = 120) (n = 173)
Less than 44 65 54.2 89 51.4
45 to 60 38 31.7 59 34.1
61 to 76 12 10.0 21 12.1
More than 76 5 4.2 4 2.3
Gender (n = 120) (n = 173)
Male 55 45.8 107 61.8
Female 65 54.2 66 38.2
Level of education (11 = 120) (n = 173)
Illiterate 17 14.2 53 30.6
Primary school 95 79.2 109 63.0
Secondary/high school 8 6.7 11 6.4
Family size (n = 120) (n = 173)
Less than 8 79 65.8 112 ’ 64.7
9 to 12 30 25.0 44 25.4
More than 13 11 9.2 17 9.8
Source of information about SC513 (n = 115) (n = 105)
Neighbors 72 62.6 3 1 29.5
Market 35 30.4 28 26.7
Extension 8 7.0 46 43.8
Source of information about NPK (n = 117) (n = 145)
Market 41 35.0 25 17.2
Extension 19 16.2 32 22.1
Neighbors 57 48.7 88 60.7
Source of information about urea (n = 118) (n = 144)
Market 41 34.7 26 18.1
Extension 19 16.1 30 20.8
Neighbors 58 49.2 88 61.1
Knowledge of advantages and
disadvantages of improved maize (n = 120) (n = 173)
varieties
Very knowledgeable 62 51.7 38 22.0
Somewhat knowledgeable 48 40.0 97 56.1
Not knowledgeable 10 8.3 38 22.0
Knowledge of application methods of _ _
NPK and urea for maize production (n — 120) (n - 173)
Very knowledgeable 49 40.8 27 15.6
Somewhat knowledgeable 45 37.5 54 31.2
Not knowledgeable 26 21.7 92 53.2

82

Table 15 shows that the majority of farmers in Machipanda (79.2 percent) and
Vanduzi (63.0 percent) had some or completed primary education, which means that they
attended or completed classes in one or more of the grades from ﬁrst grade to seventh
grade. Clearly, there were more illiterate respondents (3 0.6 percent) in Vanduzi than in
Machipanda (14.2 percent). These farmers reported that they had never attended school,
and they could not read nor write in Portuguese or Chimanhica.

To measure awareness of improved maize variety SC513, NPK and urea,
respondents were asked, “From whom did you hear about SC513?”, “From whom did
you hear about NPK?”, and “From whom did you hear about urea?” Respondents from
Machipanda and Vanduzi reported several sources of information, which were grouped
into three categories: market (seed depots, local market, informal traders, and the
neighboring country of Zimbabwe), neighbors (other farmers and relatives), and
extension services (individual or group meetings and radio broadcasts). It is important to
note that under extension, radio was the most used source of information on improved
maize variety SC513. Of the eight respondents in Machipanda who reported extension,
six (or 75 percent) mentioned radio. In Vanduzi, of the 46 respondents who reported
extension, 29 (or 68 percent) mentioned radio.

Table 15 shows that the largest percentage of farmers in Machipanda (62.6
percent) said awareness of improved maize variety SC513 came from neighbors; the
largest percentage of farmers in Vanduzi (43.8 percent) said awareness came from
extension services, so radio had an important role in disseminating information on
improved maize SC513 among farmers in Vanduzi. For respondents in Machipanda and

Vanduzi, awareness of chemical fertilizers NPK and urea came mainly from neighbors.

83

In both study areas, neighbors were equally said to be the main source of information on
NPK and urea in Machipanda (48.7 and 49.2 percent) and Vanduzi (60.7 and 61.1
percent).

In this study, respondents were also asked to self-assess their knowledge of the
advantages and disadvantages of improved maize varieties and methods of application of
NPK and urea (application of NPK at planting time and a side-dressing of urea
approximately two months after germination). To assess their knowledge, farmers were
provided with a 3-point scale, with 1 = Very Knowledgeable, 2 = Somewhat
Knowledgeable and 3 = Not Knowledgeable. The results of knowledge self-assessment
are presented in Table 15. The majority of respondents from Machipanda (51.7 percent)
felt very knowledgeable, and the majority (56.1 percent) of respondents from Vanduzi
felt somewhat knowledgeable about the advantages and disadvantages of improved maize
varieties. Advantages of improved maize varieties reported by respondents included
drought tolerance, shorter maturity time and higher production. A disadvantage reported
by respondents was the susceptibility of improved maize varieties to storage insects,
including Protesphanus truncatus and Sitophilus zeamais.

Table 15 also shows that not many farmers considered themselves knowledgeable
about chemical fertilizer application methods, particularly in Vanduzi. The majority of
respondents in Vanduzi, (53.2 percent) and a considerable number (21.7 percent) of
respondents in Machipanda, considered themselves not knowledgeable about application

methods of NPK and urea.

84

Research Objective 2

The second objective of this study was to assess farmers’ attitudes toward
improved maize variety SC513 and chemical fertilizers NPK and urea in the Machipanda
and Vanduzi administrative posts.

Attitudes are measured by asking people the extent of their agreement with
statements of belief, affect and /or behavior toward an attitude object (Eagly and
Chaiken, 1993; Aiken, 2002). For the present study, the ﬁnal attitude scales obtained
through exploratory factor analysis followed by conﬁrmatory factor analysis (as
explained in Chapter 3) revealed six statements on improved maize variety SC513, four
statements on NPK, and four statements on urea. F armers’ degree of agreement with
each statement was measured using a 5-point Likert scale with I = Strongly Agree, 2 =
Agree, 3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree. For the negative
statements, the scale was reversed to: I = Strongly Disagree, 2 = Disagree, 3 = Neutral,
4 = Agree, and 5 = Strongly Agree so that low values always indicated a favorable
attitude. This signiﬁes that a mean score of 2.3 for the negative statement “Planting
improved maize variety SC513 is waste of time and money” indicates disagreement with
the negative statement and consequently a positive attitude toward improved maize
variety SC513.

Tables 16-18 provide the percentages, mean, and standard deviation for each

attitudinal item used to measure attitudes toward SC513, NPK, and urea.

85

Farmer’s Attitudes Toward Improved Maize SC 5 I 3

Farmers were asked the extent of their agreement with six statements on
characteristics and marketability of improved maize variety SC513. The results are
presented in Table 16. In Machipanda, the highest rating was given to the statement
“Grain from SC513 is good for milling,” with a mean of 1.8 on a 5-point scale with I =
Strongly Agree, 2 = Agree, 3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree. The
majority of the respondents (93.9 percent) agreed with the statement. In Vanduzi, the
highest rating was given to the statements “Grain from SC513 is good for milling” and
“Seed of SC513 has good germination.” These two statements had mean scores of 2. 1.
The statement “Grain from SC513 is good for milling” had slightly more supporters (69.3
percent) than the statement “Seed from SC513 has good germination,” which was
supported by 67.1 percent of respondents. Very few respondents disagreed with these
two statements. About 2.2 percent disagreed with the statement “Grain from SC513 is
good for milling” and none of the respondents disagreed with the statement “Seed from
SC513 has good germination.”

In Machipanda, the second rating was given to the statements “Seed from SC513
has good germination,” with a mean score of 2.0. This statement was supported by 90.5
percent of respondents and disagreed with by 2.6 percent of the respondents. In Vanduzi,
the second rating was given to the statement “Grain from SC513 is easy to sell” with a
mean of 2.2. A considerable number of respondents (70.3 percent) agreed with the
statement, and none of the respondents (0 percent) disagreed with the statement.

In Machipanda, the third rating was given to the negative statement “When

rainfall is low local variety Chimanhica has better production than SC513,” with a mean

86

of 2.2. A considerable number of respondents (80.9 percent) disagreed with this
statement and 12 percent of respondents agreed with the statement. In Vanduzi, the third
rating was given to the statement “Planting SC513 is waste of time,” with a mean of 2.3.
Few respondents (5.5 percent) agreed with the statement, and most respondents (70.3
percent) disagreed with the statement, which indicates positive attitudes toward the
beneﬁts obtained with SC513.

In Machipanda, the fourth rating was given to the positive statement “Grain from
SC513 is easy to sell” and to the negative statement “Planting SC513 is waste of time.”
The two statements had mean scores of 2.3. About 73.9 percent of respondents were
favorable to the positive statement “Grain from SC513 is easy to sell”, and 7 percent
disagreed with the statement. The negative statement “Planting SC513 is waste of time”
received little support (13 percent). Most respondents (81.7 percent) disagreed with the
statement, which indicates positive attitudes toward the beneﬁts obtained with SC513. In
Vanduzi, the fourth rating was given to the statements “Fresh maize from SC513 is easy
to sell” and “When rainfall is low, local variety Chimanhica has better production than
SC513”. The two statements had mean scores of 2.5. The statement “When rainfall is
low, local variety Chimanhica has better production than SC513” received more support
(59.3 percent) than the statement “Fresh maize from SC513 is easy to sell,” which was
agreed upon by 52.7 percent of respondents.

In Machipanda, the ﬁfth rating was given to the positive statement “Fresh maize
from SC513 is easy to sell.” This statement had a mean score of 2.4. About 58.3 percent
of respondents agreed with the statement, and very few respondents (3.5 percent)

disagreed with the statement.

87

Farmers ’ Attitudes Toward NPK

Farmers were asked the extent of their agreement with four statements on
chemical fertilizer NPK. The responses to attitudinal statements on NPK are presented in
Table 17. In Machipanda, the highest rating was given to the positive statement “NPK
increases maize yield” with a mean score of 2. 1. About 84.6 percent of respondents
agreed with this statement, and 5.1 percent disagreed. In Vanduzi, the highest rating was
given to the positive statements “NPK is good for the maize crop” and “NPK increases
maize yield” with mean scores of 2. 1. The majority of respondents (83.4 and 85.5
percent) agreed with the statements. Very few respondents (1.4 percent) disagreed with
the statement “NPK is good for the maize crop”, and none of the respondents disagreed
with the statement “NPK increases maize yield”.

In Machipanda, the second rating was given to the positive statement “NPK is
good for the maize crop” with a mean score of 2.2. The majority of respondents (85.4
percent) agreed with the statement, and 7.7 percent disagreed. In Vanduzi, the negative
statement “NPK is a waste of time and money” received the second rating with a mean of
2.5. About 13.8 percent of respondents agreed with the statement, and more than half of
respondents (60.7 percent) disagreed with the statement, which indicates positive
attitudes toward the expenses associated with NPK. Farmers who thought NPK was not
good for maize thought so because of fertilizer burn. Farmers explained that NPK burned
their maize.

In Machipanda, the negative statement “NPK is a waste of time and money”
received the third rating with mean of 2.7. Less than half of respondents (30.8 percent)

agreed with the statement, while more than half of respondents (63.2 percent) disagreed

88

with the statement, which indicates positive attitude toward the expenses associated with
NPK. In Vanduzi, the third rating was given to the negative statement “NPK is not good
for the soil” with a mean score of 3.1. Less than half of respondents (27.6 percent)
opposed this statement, while 37.9 percent of respondents supported the statement, which
indicates that respondents had negative attitudes toward the effect of NPK on the soil.

In Machipanda, the fourth rating was given to the negative statement “NPK is not
good for the soil” with a mean score of 3.6. Few respondents (18.8 percent) opposed this
statement, while more than half of respondents (70.1 percent) supported the statement,
which indicates that respondents had negative attitudes toward the effect of NPK on the

soil.

Farmers ’ Attitudes Toward Urea

Farmers were asked the extent of their agreement with four statements on
chemical fertilizer urea. The responses to attitudinal statements on urea are presented in
Table 18. In Machipanda, the highest rating was given to the positive statements “Urea is
good for the maize crop” and “Urea increases maize yield”. The two statements had
mean scores of 1.9. The majority of respondents (97.5 percent) agreed with the statement
“Urea is good for the maize crop”, and 95.8 percent of respondents agreed with statement
“Urea increases maize yield”. In Vanduzi, the highest rating was given to the positive
statement “Urea increases maize yield”, with a mean score of 2. 1. The majority of
respondents (83.3 percent) agreed with the statement.

In Machipanda, the second rating was given to the negative statement “Urea is a
waste of time and money” with a mean of 2.2. Most respondents (85.6 percent) disagreed

with the statement, and few respondents (11.9 percent) agreed with the statement. In

89

Vanduzi, the second rating was given to the statement “Urea is good for the maize crop”
with a mean of 2.2. A considerable number of respondents (80.6 percent) agreed with the
statement, and about 1.4 percent disagreed with the statement.

In Machipanda, the third rating was given to the negative statement “Urea is not
good for the soil” with a mean of 2.4. A considerable number of respondents (72
percent) disagreed with the statement, and 17.8 percent agreed with the statement. In
Vanduzi, the third rating was given to the negative statement “Urea is a waste of time and
money” with a mean of 2.4. More than half of respondents (67.3 percent) disagreed with
the statement, and about 13.9 percent agreed with the statement.

In Vanduzi, the fourth rating was given to the negative statement “Urea is not
good for the soil” with a mean of 2.9. Unlike the negative statement “Urea is a waste of
time and money,” which about 15 percent of respondents supported, the negative
statement “Urea is not good for the soil” was supported by 31.3 percent of respondents

and less than half of respondents (35.4 percent) disagreed with the statement.

90

35888.9. 033%: 8m “883$.—

33 23m 05. SENSE bwghw u m. use 89%:qu n V .38sz u m. .meV n N 633.. 539$. u N 823 038 .9825 “509m a no woman 0.58 :82

 

 

 

 

.m 2 mom :95 5:268
w. 2 - a: ma mam 3 m. 3 - Nd o.» E E $823825.an
32 mm =ﬂﬁ8 :25,
.8585
w. mm 2 2 owm adv Tom 0. M: - 9N Wm ~60 ﬁwm woowmmmﬁmUmEobﬁﬁw
.moumﬁgow
5. mm - - 9mm Nov mdm o. 9N o. 5; OS Wan NS Bowman Emmym Eotwoom
.m N DOME nutmtmtaxgu
2233893 31832 335.2%
$288 98 08:
5. mm : :8 Nvm Que m.m w. Om n; m.: Nd v.2. me EBQBmﬂmEUmeEmE
.=8 8 >93 ﬂ mﬁmUm .«o
h. mm - #6 w—V Wow N.N 5. Wm 0. ed m.wm mém OS Amnoo waummev onmanmgm
.28
o. mm - - 5am ode Eu 5. Wm a. mo m2 m8 ms ouxmwommmEUmEBmEEO
MED”
A329. 3.85 hmELQENHS
bﬁaSﬁauE 339:8 magnum

9m. Ema X. X. X. .x. .x. QM. 33 .x. .x. .x. xx. .x.

Om Q Z < <m Dm Q 2 < <m

cans EYE
58> 88?an

E 8:03me omaommom no 3083

E 310330 omaoqmom «0 “coupon

 

 

.m 2 now 85> 8%: 8888 8.33 8835. 88283 .2 03$

91

$55283 gamma: 8,“ v8.32
33 038 of. .mm&3..5 bmzohm. u m. 38 6333.5 n V .Nuhamz u m .mmeV n N .mmkmw bmzohm u ~ 20:3 3an 3235 «£09m a no woman 20% 522

 

 

 

 

, $288
a VN V._ WQ ”a: 6% 6.5 h. Nd - Q: Wm 93 $6 98 025.3 833 x mm «ED
a QN n. 0.3 m.mm mgm m.m o. VN. - Q: N? V.V© cs d8 05 Sm team 8: mm «ED
0. _.N h. n. m.m _ WE. od m. a.“ - w. V.m can mg.— .Eoi 388 83205 «ED
0. Nd n. v; V.: 0: m6 V. a; - w. h; Nam Wm _ .388 8m voow ﬂ «ED
QM 3: .x. Ax. X. .x. AX. QM 3: X. .x. Q... .x. X.
Qm Q Z < <m Om Q 2 < <m
$3 "a a: "a
3653 3:3382

E mocowoﬁo omaoamom m0 “583 E 3:03me omaommom mo EoBom

 

 

£95 @338. mowBE< .ﬁcowqommom .M: 033.

.azoEBSm 0333: .8 3305..
mm? 038 of. $9?“me bmghm. u w EB 3&an n V .253»? u m .meV n m 633‘. A~M=ohw n N 80:3 23m BEBE “Boga a no woman 2on :82

 

92

 

 

w. 2 3 v.2 3m 3m 3 3 hm 2 EN o6 2o 2 Ba 25% 3332mm“
0. E 3.. 3m 3% mg ”a 3 9m _.: o.% a: NS 3 .=8§.5§838a§z
V. Z - - n: ma 3 c. 3 - 3 2: 3K ed .23» 3588328352
m. 3 - Z a: 3: 3 b. 3 a. ME “S no» 3 .QNRESSSXEZ
9w 3: x .x. .x. .x. .x. a. 32 .x. .x. .x. .x. x
mm a z < <m mm a z < <m
5; "3 s: "5
m§w§> 8:328:

E motowopwu omcommom mo «coach

 

 

E motowﬁmo omcommom mo 335m

.vaz Each 825? .escaamom .: 2.5

Research Objective 3

The third objective of this study was to compare attitude scores among age
categories and sources of information, and between genders of respondents within
Machipanda and Vanduzi, and to compare attitude scores between them. The results are
presented in Tables 19-22. This study sets a 5 percent level of signiﬁcance, but because

the present study is exploratory, a 10 percent signiﬁcance level was also considered.

Attitude Scores by Age Categories

Results of mean attitude by age category are given in Table 19. At 5 percent
signiﬁcance level, there were no signiﬁcant differences in mean attitude among the
following age categories: “less than 44,” “ages 45 to 60,” “ages 61 to 76” and “over 76
years” within the Machipanda and Vanduzi administrative posts. Only in Machipanda,
were there signiﬁcant (P<.10) differences in mean attitude among the age categories.
There younger farmers tended to hold a stronger positive attitude toward production

characteristics of SC513 than older farmers.

Attitude Scores by Sources of Information

Results of mean attitude by information source are given in Table 20. Sources of
information made no difference (P>.05) in mean attitude toward marketability of
improved maize variety SC513 and mean attitude toward production characteristics of
SC513 within the Machipanda and Vanduzi administrative posts.

Likewise, sources of information made no difference in attitudes toward NPK and
urea (P>.05) among respondents in Machipanda. In Vanduzi, however, there was a

signiﬁcant difference (P<.01) in mean score for attitudes toward NPK between

93

respondents who heard about NPK from neighbors and respondents who heard about
NPK ﬁ'om the extension service. Sufﬁcient evidence exists that respondents who heard
about NPK from extension tended to hold stronger positive attitudes than respondents
who heard about NPK from neighbors.

In Machipanda, there were no signiﬁcant differences in mean attitude score
among sources of information about urea. In Vanduzi, there was a signiﬁcant difference
(P<.001) in mean score for attitudes toward urea between respondents who heard about
urea from neighbors and respondents who heard about urea from extension. As with
responses about NPK sufﬁcient evidence exists that respondents who heard about urea
from extension tended to hold stronger positive attitudes than respondents who heard

about urea from neighbors.

Attitude Scores by Gender of Respondent

Results of mean attitude by gender of respondents are given in Table 21. In
attitudes toward marketability and production characteristics of SC513, the gender of
respondents made no difference (P>.05) in Machipanda and Vanduzi. For attitudes
toward NPK, the gender of respondents did not make a difference in Machipanda but did
make a difference in Vanduzi (P<.05), with male respondents holding stronger positive
attitudes than female respondents. For attitudes toward urea, the gender of respondents
made a difference (P _<_ .01) in Machipanda and Vanduzi. In both study locations, male

respondents tended to hold stronger positive attitudes than female respondents.

94

 

 

 

 

 

 

 

 

3.Vm ...

 

 

 

 

2: Q: m 5mm. omd m 3. 55 222
mm: mmd cm on; No.5 3 3. 9 5
3m. no; SN nmd mm m 3 n. cos. amd om.” cm co 9 Q
_N.N and am mm; mm.w mo 3 55 $04 .88 @338 £6333
Rm. m.3 m 3N v.3 m 3. $5 262
am; Rd 3 3w.— oo.: 3 on 8 5
com. mm; $3 mud E C _. SN NmN 3.3 cm cc 9 me
2: 3.3 mm mm; 3.3 3 3 55 $3 .vEZ 2558 8353
no; oﬁw v :3 cos n on 55 202
$3 $6 2 mm; $6 2 on 8 G .mﬁmom mo
wmo. new. :.N :6 em moo. ...3.N 33 new om ow 8 we 8538835 :28:on
N5; 056 Nv 3m." m ﬁe mo 3 55 $3 @338 3552
cc; one v 2: 86 m on as: 202 .3 wow
No.3 8.5 S mmd cod 2 on 9 S burg 3.38 35383
mum. 3m. mm; ems 1m m3. ow; mm; $6 on co 9 me («0 5:388.wa
mm; 3.0 NV 33 3.0 no 3 52: v.84 @339 368:2
mm m a. 52 2 ma a an :82 z we.
“won oZHngEvm 365$ “mom o>uabm§8um 3:89:82

stowage owe. 3 03?. as: do mares. $3-80 é 03¢ <>oz< .2 as“;

 

 

95

 

 

Sovm .I. SVm ... 60508 cemtmmEoo 23:2: 8: “won Pomonom

 

 

 

 

 

 

 

 

 

8N 8.2 3 as no.” mm 32585302
25. m3 SN m3 8 SN. 42 m: a; 2 3:29:53
”3 3d om of m3 :. Ba: .85 case 033.
:2 2: 3 m3 3: a ...esoﬁwez
.25. Sun a: 8.0 mm o8. as. a: 2: 2 2280me
N: Nos 8 a 3 2: :. Ea: .vaz 333 $38...
2: is R 2: m3 w soacoim
So. a; a: 2: 3 as. so. 3: 8.0 mm Ea:
m: 2: wm a: 9% Q 2.8902 .28m do 855836
GomuOSVOHQ 6.830..— OUBE<
N2 MS 2.. :._ as w :28me .38
9:. 2d ”2 m: 8 ME. SN N2 as mm 332 35> use Ezaeco
o? 2; mm #2 :3 Q $8302 $59335 Easggg
mm a an as: 2 mm a ma. 502 z 8928an
880m
“mom 03353583 ﬁscqm> “mom gumbﬂﬁeg «983522

 

 

33:8an 32:8 3 02:3 502 do gage .8320 5 22¢ <>oz< .8 use.

96

oﬁvm ... movm .1. EVA“ 3.2.,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

wad— ov Sum we 2&th

coo. ...,...owN NS. Iomd
Rd ma 2:. Vm 032 .85 @839 235:2
m2: 3 5.3 No oEEom

Duo. ICON . SN. 3 ._ .
me a 3 cm 3 mm 232 .vEZ @339 owBE<
v2. 3 omd no oREom

no _. om; . 23. *2; . .m ~ mom mo 85:808an
Vm w on S m mm 232 5326qu @338 0632?.
3.5 E w ﬂ .5 no 03.5““

mg. 3w. . 2:. mo; . .m _ mum 53:3 SEE 32:95
8 b E 8 0 mm 2“: we €398.35 Baas $35....

337% 333 E 2 mégi 3:93 E 2 8280
“won 0235383 365$ “mom 0235??? 35928.2

H
H

.ﬁcovcoamom go 8280 3 82:2 :82 AN 2an

97

Attitude Scores by Study Location

Results of mean attitude by study location are given in Table 22.

Table 22. Mean Attitudes by Study Location.

 

 

 

 

Study Location N M t-value P-value

Attitude toward marketability of Machipanda 115 6 94
improved maize variety SC513. ° .568 .571

Vanduzi 91 7.07

Attitude toward production Machipanda 115 6 05
characteristics of SC513. ' 2.79“ .006

Vanduzi 91 6.68

Attitude toward NPK. Machipanda

“7 10'6 3.19* .002

Vanduzi 145 9.83

Attitude toward urea. Machipanda 118 8.32
- 5.21 ** .000

Vanduzi 144 9.59

*P<.01 **P<.001

In attitudes toward marketability of SC513, the study location made no difference

(P>.05). However, signiﬁcant differences (P<.01 and P<.001) in mean attitude scores

were observed between Machipanda and Vanduzi, in attitudes toward production

characteristics of SC513, attitudes toward NPK, and attitudes toward urea. Respondents

in Machipanda tended to hold stronger positive attitudes toward production

characteristics of SC513 than respondents in Vanduzi. Respondents in Vanduzi tended to

hold stronger positive attitudes toward NPK than respondents in Machipanda.

Respondents in Machipanda tended to hold stronger positive attitude toward urea than

respondents in Vanduzi.

98

Research Objective 4

The fourth objective of this study was to describe the pattern of adoption of
improved maize variety SC513, NPK and urea in Machipanda and Vanduzi

administrative posts.

Pattern of Adoption of Improved Maize Variety SC 5 1 3, NPK and Urea

Figures 9, 10 and 11 show the pattern of adoption of SC513, NPK and urea in
Machipanda and Vanduzi. The curves are based on respondents’ recall (Byerlee and de
Polanco, 1986) of the year in which they ﬁrst used the technology. Figure 9 presents the
pattern of adoption of SC513. In the high region (Machipanda), the percent of usage
increased slowly between 1992 and 1994 from 9 percent to 13 percent. For the same
time period (1992-1994), the percent usage was almost constant at very low rates (1
percent) in the low region (Vanduzi). From 1995 to 2004 in the high region
(Machipanda), the percent usage increased rapidly, and by 2000, about 45 percent of
farmers in Machipanda had used hybrid maize variety SC513. For the same period
(1995-2004), an increase in the rate of adoption, from about 1 percent to 28 percent of
adopters, occurred in the low region (V anduzi). Machipanda reached about 88 percent
usage by 2004; Vanduzi reached a lower percent usage (about 28 percent).

Figure 10 shows the pattern of adoption of NPK. The adoption pattern for NPK
was also different between the high region (Machipanda) and the low region (V anduzi),
with the high region reaching a higher percent usage more quickly than the low region.
By 2005, Machipanda reached almost 34 percent usage, and Vanduzi reached
approximately 16 percent. By 1995, both study areas were experiencing a steady

increase in farmers applying NPK on maize.

99

 

100

80

  
  
 

60 Machipanda

40
Vanduzi

20

/s+--- -

.. -»~?’ ’4

 

 

0 o—- ———{+—— vir— - 43'

 

Cumulative percentage of farmers adopting SC513

1992 1995 1998 2001 2004
Year started cultivating hybrid maize SC513

 

Figure 9. Pattern of Adoption of Hybrid Maize SC513 in Machipanda and Vanduzi from
1990 to 2005.

 

40

30
Machipanda

20
Vanduzi

.}+ Q“

10

 

cf

0 f»4)~-—- —— ——-—~—- 43-0”

 

 

 

Cumulative percentage of farmers adopting NPK

1975 1980 1985 1990 1995 2000 2005
Year started applying NPK on maize

Figure 10. Pattern of Adoption of Fertilizer NPK in Machipanda and Vanduzi from 1977
to 2005.

100 l

Figure 11 shows the pattern of adoption of urea. Like the patterns of adoption of
SC513 and NPK, the pattern of adoption of urea differed between the study areas, with
the high region reaching a higher percentage of usage more quickly than the low region.
By 2005, Machipanda had reached approximately 53 percent usage, while Vanduzi
reached approximately 17 percent. From 1995 on, both study areas had experienced a

steady increase in the number of farmers applying urea on maize.

 

60

 

50

Machipanda
40

30
Vanduzi

20

10

 

 

 

 

 

Cumulative percentage of farmers adopting urea

1970 1980 1990 2000 2010

Year started applying urea on maize

Figure 11. Pattern of Adoption of Fertilizer Urea in Machipanda and Vanduzi from 1972
to 2005.

The three patterns of adoption in Figures 9, 10, and 11 show that, from 1995,
when the improved maize SC513 and chemical fertilizers NPK and urea were formally
introduced by the DNER/SG2000 extension program, the rate of adoption increased for

all three technologies.

101

Adoption of Improved Maize Variety SC 5 1 3 and Fertilizers NPK and Urea in the last two

years (2004-2005)

Very few farmers, particularly in the lowlands of Vanduzi, adopted continually
improved maize variety SC513 and chemical fertilizers NPK and urea in the last two
years (2004 and 2005). In Machipanda, 45 out of 120 respondents (37.5 percent)
reported cultivation of SC513; in Vanduzi, only 10 out of 173 respondents (5.8 percent)
reported cultivation of SC513. Regarding use of chemical fertilizer (NPK/and or urea) in
Machipanda, 39 out of 117 respondents (33 percent), reported application of chemical
fertilizer on maize. In Vanduzi, only 4 out of 145 respondents (3 percent) reported

application of chemical fertilizer on maize.

Discontinuance of SC 5 1 3, NPK and Urea

The current (2004 and 2005) adoption indicates that discontinuance occurred

throughout the time since the ﬁrst years farmers started using the technologies (Table 23).

Table 23. Percentage of Farmers Who Ever Used and are Currently Using the
Technology.

 

 

 

Technologies Study location
Machipanda Administrative Vanduzi Administrative
Post Post
Ever used Using 2004-05 Ever used Using 2004-05
Hybrid maize SC 513 88% (105) 38% (45) 28% (48) 6% (10)
Fertilizers (NPK/urea) 87% (102) 33% (39) 33% (48) 3% (4)

For example, in Machipanda, 50.5 percent of respondents and in Vanduzi, 22.2
percent discontinued cultivation of SC513. In Machipanda, 2 percent of respondents and

in Vanduzi, 15 percent discontinued application of NPK on maize. In Machipanda, 20

102

percent of respondents and in Vanduzi, 13.8 percent discontinued application of urea on

maize.

Reasons for Discontinuance of SC 51 3, NPK, and Urea

The reasons for discontinuance of SC513, NPK, and urea were also explored.

Table 24 shows the reasons for discontinuance of hybrid maize SC513.

Table 24. Reasons for Discontinuance of Use of Hybrid Maize SC513 Among Farmers in
Machipanda and Vanduzi.

 

 

Machipanda Vanduzi
Administrative Post Administraive Post
Frequency Percent Frequency Percent

 

Reasons for discontinuance

 

Seed availability 4 10.8 1 2.9
Difﬁcult to save seed 1 2.7 3 8.6
Storability 2 5.4 9 25.7
Husk cover 16 43.2 2 5.7
Lack money 8 21.6 7 20.0
Other (better varieties) 6 16.2 13 37.2
Total 37 100 35 100

 

Farmers reported several reasons for discontinuing use of SC513. In Machipanda,
husk cover was the most reported (43.2 percent) reason for discontinuance of SC513.
Farmers explained that the husks of SC513 did not close well, which exposed the cob to
rain and attack by ﬁeld insects and diseases (fungus). Lack of money to purchase SC513
was the second most (21.6 percent) reported reason for discontinuance of SC513, and
preference for other hybrids (SC700) considered better than SC513 was the third most
(16.2 percent) reported reason for discontinuance of SC513 in Machipanda. In Vanduzi,
preference for better varieties (PAN, Matuba, Chimanhica, Candjere) was the most
reported reason (3 7.2 percent) for discontinuance of SC513. Storability of SC513 was

the second most (25.7 percent) reported reason for discontinuance of SC513, and lack of

103

money to purchase SC513 was the third most (20 percent) reported reason for
discontinuance of SC513. Tables 25 and 26 show the reasons for discontinuance of
chemical fertilizers NPK and urea. Among several reasons for discontinuing use of NPK
and urea, lack of money to purchase fertilizers was the most reported reason for

discontinuing use of NPK and urea in Machipand and Vanduzi areas.

Table 25. Reasons for Discontinuance of NPK Use Among Farmers in Machipanda and

 

 

 

Vanduzi.
Machipanda Vanduzi

. . Administrative Post Administraive Post
Reasons for drscontrnuance Frequency Percent Frequency Percent
Lack of credit from DNER/$62000 - - 3 13.6
Lack money to purchase NPK 20 62.5 16 72.7
Plant local variety 2 6.25 1 4.55
Soil is fertile 2 6.25 2 9.09
Damage soil 8 25 - -
Total 32 100 22 100

Table 26. Reasons for Discontinuance of Urea Use Among Farmers in Machipanda and

 

 

Vanduzi.
Machipanda Vanduzi

. . Administrative Post Administraive Post
Reasons for discontinuance Frequency Percent Frequency Percent
Lack of credit from DNER/SG2000 - - 3 12.5
Lack money to purchase urea 23 71.9 17 70.8
Plant local variety 1 3. 13 1 4.17
Soil is fertile 3 9.38 2 8.33
Damage soil 3 9.38 - -
Other (lot of work, drought) 2 6.25 1 4.17
Total 32 100 24 100

Machipanda and Vanduzi differed markedly in two reasons for discontinuance of
NPK and urea: “Lack of credit from DNER/$02000” and “Damage soil”. The ﬁrst was
reported only by respondents in Vanduzi, and the second was reported only by

respondents in Machipanda.

104

 

Research Objective 5

The ﬁfth objective of this study was to determine the factors associated with
adoption of improved maize variety SC513 and NPK and urea chemical fertilizers in the
Machipanda and Vanduzi administrative posts.

In conformity with the theoretical framework illustrated in Figure 6 in Chapter 2,
four attitudinal dimensions - “attitudes toward production characteristics of SC513,”
“attitudes toward income from SC513,” “attitudes toward effects of NPK and urea on
maize,” and “attitudes toward costs of NPK and urea,” - were hypothesized to capture the
variable farmers’ attitudes toward improved maize variety SC513, NPK, and urea. The
ﬁnal attitudinal scales were obtained through Exploratory and Conﬁrmatory Factor
Analysis and were presented in Chapter 3. To include the four attitudinal dimensions as
explanatory variables of adoption, a two-step procedure was followed. First, a factor
analysis was performed to estimate the attitude scores. Second, a multiple regression
model was performed to determine the factors associated with adoption. The estimated
attitude scores were included in the regression as explanatory variables of adoption. The
two steps are described below.

Step 1: Factor Analysis

To create attitude scores for subsequent use in logistic regression, factor scores
were estimated through the principal component and varimax procedure (Hair et al.,
2005). High factor loadings were obtained. Two-factor solutions for the attitude scale on
SC513 explained about 59.3 percent of the variance in attitudes toward SC513. A one-
factor solution for the attitude scale on NPK explained about 48 percent of the variance in

attitude toward it. A one-factor solution for the attitude scale on urea explained about

105

 

56.4 percent of the variance in attitudes toward urea. The values of percent of variance
explained that were obtained in this study are acceptable in the social sciences (Hair et
al., 2005). The reason for using the varimax procedure is that it does not assume
correlation between factors.- This is advantageous because the researcher can avoid
complications caused by multicollinearity (Hair et al., 2005).

Step 2: Multiple Logistic Regression

 

A multiple logistic regression model was used to determine the factors associated
with adoption of improved maize variety SC513. The dependent variables were adoption
of improved maize variety SC513, adoption of the chemical fertilizer NPK, and adoption
of the chemical fertilizer urea.

The independent variables were agro-ecological zone (AEZ), age (AGE), family
size (F AM), education (EDU), how-to knowledge (KNOW), neighbors (N EIG), market
(MARK), extension (EXT), attitudes toward marketability of SC513 (ATTmark), attitude
toward characteristics of improved maize SC513 (ATTchar), attitude toward
NPK(ATTnpk) and attitude toward urea (ATTurea) on adoption of SC513, NPK, and
urea in the high region (Machipanda) and the low region (V anduzi). Tables 27-29 show
the maximum likelihood estimates of logistic models. Each table is followed by a brief

interpretation of the statistically signiﬁcant empirical logistics estimates.

Factors Associated with Adoption of Improved Maize SC513

The logistic model representing the factors associated with adoption of improved
maize variety SC513 is

Log (rt/1 — 1:) = or + BIAEZ + |32AGE + B3FAM + B4EDU + BsKNOW + BNNEIG +
BMMARK + BGATTmark + B7AT'Tchar + BgADP*ATTmark

106

 

Table 27 presents the maximum likelihood estimates of the logistic model for
factors associated with adoption of SC513 among respondents in the Machipanda and
Vanduzi administrative posts. The ﬁt of the model is satisfactory. The estimated
coefﬁcient for the likelihood ratio chi-square = 87.03 and is signiﬁcant (P<.001). The

model accounts for about 39 percent of the variation between adopters and non-adopters
of SC513 (R2 Logistic = 0.388) (Appendix F). For logistic regression, lower percentages

of variance are observed (Hosmer and Lemeshow, 2000).

Table 27. Maximum Likelihood Estimates of the Logistic Model for Factors Affecting
Adoption of SC513 Among Respondents in Machipanda and Vanduzi.

 

Variables B S.E. Wald Sig. Exp([3)
Agro-ecological zone 2.51 .512 24.0 .000 12.2
Age (years) -.020 .014 2.18 .140 .980
Family size .005 .063 .008 .930 1.01
Level of education .745 .566 1.73 .189 2.11
Knowledge of advantages and
disadvantages of S C 513 1.99 .732 7.40 .007 7.32
Information sources on SC513 3.47 .176
Neighbors (1) -.745 .604 1.52 .218 .475
Market (2) -1.15 .618 3.46 .063 .317
Attitude toward marketability of produce
from S C 51 3 -.497 .234 4.50 .034 .608
Attitude toward traits of SC513 -.812 .228 12.6 .000 .444
* . . .
ggrsnllgost Attrtude marketability of .815 '472 2.98 .084 2.26
Constant -.704 1.33 .279 .598 .495
Likelihood ratio chi-square of (10) 87. 03 .000

df' degree of freedom for the model. '

Interpretation of the Coeﬂicients
Among the factors considered in the model, ﬁve were found to have a signiﬁcant
effect on the adoption of improved maize variety SC513. In order of magnitude of effect,

these factors are agro-ecological zone, attitudes toward traits of SC513, knowledge of

107

 

advantages and disadvantages of SC513, attitudes toward marketability of produce from
SC513, and interaction between agro-ecological zone and farmers’attitude toward
marketability of SC513 (AEZ*Attitude marketability SC513). The logistic regression
coefficients are interpreted using values of odds ration Exp ([3) because it can provide the
amount of increase in probability associated with the independent variable.

As expected, the agro-ecological zone and the knowledge of advantages and
disadvantages of SC513 had signiﬁcant positive effects on the adoption of improved
maize variety SC513 at 1 percent signiﬁcance level. The probability that farmers in the
highlands of Machipanda cultivated hybrid maize SC513 in the period 1995-2005,
increased by a factor of 12, indicating that farmers in Machipanda are more likely to
adopt hybrid maize SC513 than farmers in the lowlands of Vanduzi. Similarly, the
probability that farmers who felt knowledgeable about the advantages and disadvantages
of improved maize varieties cultivated improved maize variety SC513 increased by a
factor of 7, indicating that knowledgeable farmers are more likely to adopt improved
maize variety SC513 than non-knowledgeable farmers.

As expected, holding negative attitudes toward traits of SC513 and negative
attitudes toward the marketability of produce from SC513 had a signiﬁcant negative
effect on adoption of it at 1 and 5 percent signiﬁcance levels. An increase of one point in
the scale of “Attitudes toward production traits of improved maize variety SC513”, which
indicated an unfavorable attitude, decreased the logarithm of the odds of adoption of it by
.812, and an increase of one point in the scale of “Attitudes toward marketability of
maize variety SC513” decreased the logarithm of the odds of adoption of the variety by

.497. The odds ratio Exp ([3) less than 1 indicates that the probability of adoption for

108

 

 

farmers with negative attitudes is lower than the probability of adoption for farmers with
positive attitudes.

For the interaction term, AZE*Attitudes toward the marketability of SC513, the
coefﬁcient ([3 = .815) indicates that the difference between the rate of change in Vanduzi
and Machipanda increased as the scores on attitudes toward marketability of produce
from maize SC513 increased (or as the attitude toward marketability became negative).
In other words, as the attitudes of farmers in Machipanda tended to be negative by one
score, farmers in Machipanda were about 2 times more likely to adopt the hybrid SC513

than farmers with negative attitudes in Vanduzi. Other variables in the model (age, level

 

of education, family size and sources of information) were not signiﬁcant.

Factors Associated with Adoption of NPK Fertilizer

The logistic model representing the factors associated with adoption of NPK is

Log (it/l - 71:) = (I. + B] AEZ + BZAGE + B3FAM + B4EDU + BSKNOW + BMMARK +
BgEXT ‘l' B6ATTnpk

Table 28 presents the maximum likelihood estimates of the logistic model for
factors associated with adoption of NPK among respondents in the Machipanda and

Vanduzi administrative posts.

109

 

Table 28. Maximum Likelihood Estimates of the Logistic Model for Factors Affecting
Adoption of NPK Among Respondents in Machipanda and Vanduzi.

[3 SE. Wald Sig Exp([3)

 

Agro-ecological zone .794 .392 4.10 .043 2.21
Age (years) -.011 .013 .680 .410 .989
Family size .041 .052 .632 .427 1.04
Level of education .246 .552 .199 .656 1.28
Knowledge of application methods for NPK 2.61 .753 12.0 .001 13.6
Information sources on NPK 14.5 .001

Market (1) .154 .442 .122 .727 1.17

Extension (2) 1.57 .430 13.3 .000 4.81
Attitude toward NPK -.322 .208 2.40 .122 .724
Constant -4.32 1.11 15.1 .000 .013
Likelihood ratio chi-square df (8) 64.3 .000

 

dfl degree of freedom for the model.

The ﬁt of the model is satisfactory. The estimated coefﬁcient for the likelihood

ratio chi-square = 64.3 and is signiﬁcant (P<.001). The model accounts for about 24
percent of variance in adoption of NPK (R2 Logistic = 0.243) (Appendix F).

Interpretation of the Coeﬂicients

Among the factors considered in the model, three were found to have a signiﬁcant
effect on the adoption of NPK. In order of magnitude of effect, these factors are
information sources on NPK, knowledge of application methods for NPK, and agro-
ecological zone. As expected, the extension services as sources of information on NPK,
the farmers’ knowledge about the methods of application of NPK, and the agro-
ecological zone had signiﬁcant positive effects on the adoption of NPK, at 1 and 5
percent signiﬁcance levels. Farmers who learned about NPK from extension services had
a probability of adoption of NPK about 5 times the probability of adoption of NPK of
farmers who learned about NPK from neighbors, indicating that farmers who use

extension services are more likely to adopt NPK than farmers who leaned about NPK

110

from neighbors. Farmers who were knowledgeable about the methods of application of
NPK were more likely to adopt NPK by a factor of 14. Likewise, the probability that
farmers in the highlands of Machipanda applied NPK on maize increased by a factor of 2,
indicating that farmers in Machipanda were more likely to adopt NPK for maize
production than farmers in the lowlands of Vanduzi. Other variables in the model (age,
level of education, family size, and attitudes toward chemical fertilizer NPK) were not

signiﬁcant.

Factors Associated with Adoption of Urea Fertilizer

The logistic model representing the factors associated with adoption of urea

 

fertilizer is
Log (TI/1 - 1!) = (I 'l' ﬁlAEZ 'l' BzAGE + B3FAM + B4EDU 'l" B5KNOW + BMMARK +
BEEXT + B6ATI‘urea

Table 29 presents the maximum likelihood estimates of the logistic model for
factors associated with adoption of urea among respondents in the Machipanda and

Vanduzi administrative posts.

Table 29. Maximum Likelihood Estimates of the Logistic Model for Factors Affecting
Adoption of urea Among Respondents in Machipanda and Vanduzi.

 

Variables [3 SE. Wald Sig. Exp(B)
Agro-ecological zone 1 .43 .369 14.9 .000 4.16
Age (years) -.013 .012 1.13 .288 .987
Family size -.018 .050 .133 .716 .982
Level of education .256 .525 .238 .626 1.29
Knowledge of application methods for urea 2.05 .521 15.5 .000 7.74
Information sources on urea 9.99 .007

Market (1) -.211 .418 .256 .613 .810
Extension (2) 1.23 .434 8.01 .005 3.41
Attitudes toward urea -.251 .203 1.53 .216 .778
Constant -2.86 .964 8.81 .003 .057
Likelihood ratio chi-square df (8) 81.6 .000

di' degree of freedom for the model.

111

 

The ﬁt of the model is satisfactory. The estimated coefﬁcient for the likelihood

ratio chi-square = 81.6 and is signiﬁcant (P<.001). The model accounts for about 27

percent of variance in adoption of urea (R2 Logistic = 0.272) (Appendix F).

Interpretation ofthe Coefficients

Among the factors considered in the model, three were found to have a signiﬁcant
effect on the adoption of urea fertilizer. In the order of magnitude of effect, these factors
are knowledge of application methods for urea, agro-ecological zone, and information

sources on urea.

 

As with adoption of NPK, the adoption of urea was positively inﬂuenced by
knowledge about the methods of application of urea, the agro-ecological zone, and the
extension services as sources of information on urea at 1 percent signiﬁcance level. For
knowledgeable farmers, the probability of adoption of urea was about 8 times the
probability of adoption of urea for non-knowledgeable farmers. Likewise, for farmers in
the highlands of Machipanda, the probability of adoption of urea was about 4 times the
probability of adoption of urea for farmers in the lowlands of Vanduzi. For farmers who
learned about urea from extension services, the probability of adoption of urea was about
3 times the probability of adoption of urea by farmers who learned about urea from
neighbors. Other variables in the model (age, level of education, family size, and attitude

toward urea) were not signiﬁcant.

112

 

CHAPTER V

DISCUSSION

This chapter presents the researcher’s interpretations of the results of the study.
The interpretations are based upon the literature, ﬁeld observations and researcher’s
knowledge, feelings, and personal experiences gained during this investigation.
Characteristics of Respondents

Maize growers from the highlands of Machipanda and the lowlands of Vanduzi
were fairly similar in age and family size. The average age of respondents in
Machipanda and Vanduzi was about 44 years, and the average family size was about 7.8
members per household in Vanduzi and 7.6 in Machipanda. Other survey research in
Mozambique and elsewhere had reported similar results on the average age of farmers
(Trabalho the Inquérito Agricola [TIA] (Agricultural Survey), 2005; Kaliba et al., 2000).
The present results on average family size are slightly higher than those found in previous
research in rural Mozambique (TIA, 2005), which indicated an average of around six
members per household.

Theoretically, the younger the farmers, the more likely they are to adopt a new
technology than older farmers because they have had more schooling (CIMMYT, 1993),
and they are more susceptible to attitude change (Visser and Krosnick, 1998). The
majority of farmers in the study areas have an age range from 19 to 43 years, so chances
are that a proportion of the population targeted by the public extension is highly
susceptible to attitude change and eventually adopted the improved maize varieties and
chemical fertilizers. Assuming that adoption of new varieties requires more labor inputs

(Feder et al., 1985), we would think that the rural households in Machipanda and

113

Vanduzi have relatively large (i.e., above average size) families to rely upon for labor. It
is important to note, however, that the total number of family members does not always
mean availability of labor because there must be families with higher dependency rates
than others (i.e. families with more children to take care of). Though the relatively large
family size may indicate more labor for cultivation, more precise results on labor
availability would require information on dependency ratios.

The distributions of age and family size (Table 15) were very similar in
Machipanda and Vanduzi administrative posts. However, this was not the case for other
characteristics - namely, gender, level of education, sources of information on improved
maize variety SC513 and chemical fertilizers NPK and urea, knowledge about advantages
and disadvantages of improved maize varieties, and knowledge about application
methods of NPK and urea. Machipanda had more women respondents than Vanduzi.
There are several possible explanations for the difference in gender of the respondents.
One obvious explanation is that, in Machipanda, male farmers were absent during most
of the ﬁeldwork. It is important to note that about maize cultivation both male and
female respondents considered themselves knowledgeable. Thus, despite the fact that
Machipanda and Vanduzi follow a patrilineal system, whereby the man is the head of the
household and this gives him the right to talk with outsiders, in the absence of the men
women were interviewed. This happened frequently in Machipanda, where during the
ﬁeldwork, men were not present, or if they were present, they were sick and not able to
answer the questions. On the other hand, the men in Vanduzi were mostly available for
the interviews - i.e., they, were present at the time of the interviews and qualiﬁed

themselves as knowledgeable about maize cultivation. The second explanation is

114

somewhat less obvious and is based on the differences between Machipanda and
Vanduzi, in land use and the main uses of the maize produced, which may play a role in
males’ level of commitment in maize cultivation. In Machipanda, maize output is mainly
used for home consumption. This may explain men being less involved with maize
cultivation in Machipanda than in Vanduzi, where maize output is also mainly used for
sale and where men are more involved with maize cultivation. Another possible
explanation is that, the interviews were conducted during the harvesting period, and
maize harvesting is usually done by women, so women were more likely to have been
available (in Machipanda) while men were more involved in other income-generating
activities. In fact, women in Machipanda, when asked about their husbands, often replied
that the men were involved in small business, such as going back and forth to the
bordering country of Zimbabwe to sell and buy consumable goods (other than maize).

On the other hand, in the remote villages of Vanduzi, men could often be seen
transporting and storing maize, and performing a major role of a middleman who
consolidates maize grain.

Shifting cultivation in Vanduzi may demand more male labor than maize
cultivation in Machipanda. In Vanduzi, I interviewed some male farmers who had just
ﬁnished opening new lands for agriculture. This indicates that farmers in Vanduzi
experienced certain freedom to practice shifting cultivation. Moreover, the opening of
new and dense areas is done manually, and it is an activity that is usually performed by
male farmers. On the other hand, in Machipanda, although farmers considered shifting
cultivation a common agricultural practice, this practice was constrained by forest

plantation programs. Indeed, during the interviews, some farmers in Machipanda took

115

the opportunity to explain how productive they thought the lands under the forest were,
but forest guards were not allowing farmers to use those lands for agriculture. Thus, with
almost no shifting cultivation in Machipanda, male labor for the family farm might have
not been needed as much as in Vanduzi.

The ﬁndings on level of education suggested that more farmers in Machipanda
attended primary school and thus were more in a position to read and write in Portuguese
and possibly perform basic computation skills than farmers in Vanduzi. Lack of access
to schools and having served in the army are some of the possible explanations why the
majority of illiterate respondents were found in Vanduzi. Several respondents in
Machipanda reported attending school in the neighboring country of Zimbabwe, a
country that a few years ago was doing well economically. In Vanduzi, several farmers
were former soldiers who fought in the ﬁfteen-year civil war, which ended in 1992.
There is a possibility that these soldiers went into the army while were illiterate and did
not go to school after they got out of the army.

Low levels of schooling have also been found among Tanzanian farmers (Kaliba
et al., 2003). One possible problem with a low level of education is that it may constrain
farmers’ ability to perceive changes in economic conditions (such as relative prices) and
to respond efﬁciently. Moreover, formal education and extension services are crucial to
predispose farmers to take an interest in a new technology and make informed decisions
about the technology. Being able to read in Portuguese or the local language
(Chimanhica) is crucial for the dissemination of improved seed among farmers in
Machipanda and Vanduzi, mainly because the packs containing seed also have basic

written instructions on crop management, including seed handling, planting, and weed

116

 

control. Lack of formal education is commonly considered more of a constraint for the
adoption of chemical fertilizer than for the adoption of improved maize seed because
chemical fertilizers are complex technologies which require some basic computation
skills for fertilizer rates (CIMMYT, 1993). However, if computation skills are to be
addressed by formal education, we also have to be cautious not to neglect the ad0ption of
improved maize, because recommendations on improved maize also include aspects of
computations, such as plant densities and their corresponding seed rates. Extension
educational programs may sufﬁce (in certain cases they may, to some extent, replace
formal education) to raise awareness and provide farmers with how-to knowledge about
cultivation of improved maize seed and application of chemical fertilizer, but if
computaions skills are required and the extension effort is equally applied to illiterate
farmers as to farmers with some education, chances are great that only farmers with some
schooling will be in a good position to understand and implement correctly the extension
recommendations for both cultivation of improved maize and application of chemical
fertilizer. Thus, intervention aimed at improving the level of education, particularly of
illiterate farmers, should be considered.

As sources of information, neighbors and the public extension service were the
most reported sources of information on improved maize variety SC513 and chemical
fertilizers NPK and urea among farmers in Machipanda and Vanduzi. As sources of
information, neighbors and extension are important when disseminating new
technologies, particularly because both sources (neighbors and extension) involve
personal communication, an important means to inform and persuade farmers to use the

innovative technologies (Rogers, 1995; Ryan and Gross, 1943). Hence, farmers in

117

Machipanda and Vanduzi are mostly using (and should continue using) those means of
communication that have a high chance of being effective in persuading farmers to adopt
improved maize varieties and chemical fertilizers. It is important to note however, that,
for dissemination of improved agricultural technology, extension should be given priority
because extension messages are planned to ensure persuasive communication, which is
probably the most common way of trying to change person’s attitudes (Oskamp and
Schultz, 2005). Neighbors may lack established formal means for the diffusion of
innovation and thus by chance reach small number of farmers. Moreover, within the
agricultural sector, it is the extension’s responsibility to communicate with farmers
through a range of methods, including radio, result demonstration plots, and farmers’
ﬁeld days, to create awareness of improved technology. Within extension, radio was the
most used information source to form beliefs about improved maize variety SC513 in
Machipanda and Vanduzi. Radio broadcasting included general information on brand
name, prices, and retailers of improved maize varieties. In Mozambique, radio
broadcasting has been effective in raising farmers’ awareness about improved maize seed
(Massingue et al., 2004). Other studies have also shown that mass media are more used
by farmers to learn about general information regarding the technology (Lichtenberg and
Zimmerman, 1999). Thus, both radio broadcasting on improved seed and farmers’ access
to radio should be encouraged.

Maize growers in Machipanda and Vanduzi were knowledgeable of the
advantages and disadvantages of improved maize varieties. On several occasions, when
respondents were asked to assess how knowledgeable they were on the advantages and

disadvantages of improved maize varieties, they explained that improved maize varieties

118

have the advantage of coping with drought and securing a food supply for their
households. Research in southern Africa (Byerlee, 1994) also pointed out that farm
households sometimes seek to improve food security by planting an earlier maturing
variety that can be consumed in the “hungry season” before the main harvest. One
disadvantage of improved maize that respondents in this study mentioned was the
susceptibility of the improved maize variety to infestations of maize weevil when stored
unshelled with the husks intact (the common way maize is stored). As mentioned in the
previous chapters, when farmers incur postharvest losses due to storage weevils
(including protesphanus truncatus and Sitophilus zeamais), they tend to sell their produce
when prices are low, therefore losing money. To help control postharvest losses,
extension and research should collaborate to develop maize varieties with more resistance
to storage insects and/or promote alternative storage facilities that can minimize insect
damage.

Though the knowledge about the advantages and disadvantages of improved
maize varieties appeared to be well established in both study locations, this was not the
case for knowledge about application methods of NPK and urea fertilizers on maize. The
majority of respondents (53.2 percent) in the lowlands of Vanduzi and a considerable
number of respondents (21.7 percent) in Machipanda considered themselves as lacking
knowledge about application methods of NPK and urea. One of the problems that
extension services may face when farmers lack knowledge on how to use an innovation,
is the rejection of the technology by the farmers (Rogers, 1995). This implies that, to
ensure adoption of chemical fertilizers, farmers have to feel knowledgeable about the

methods of application. Therefore, the National Extension Directorate (DNER) should

119

collaborate with specialists in soil fertility management to design training programs on
the optimal use of fertilizer for farmers, particularly in the lowlands of Vanduzi.
Farmers ’Attitudes Toward Improved Maize Variety SC 5 I 3 and NPK and Urea
Fertilizers

In general, respondents from Machipanda and Vanduzi held positive attitudes
toward the marketability of improved maize variety SC513, the production characteristics
of SC513, and use of NPK and urea for maize production. For improved maize variety
SC513, most farmers in Machipanda and Vanduzi appreciated the quality of the seed and
maize meal obtained with maize variety SC513. Other studies found that farmers are
favorable toward the production characteristics of improved maize varieties (Hintze et al.,
2003). Farmers in Vanduzi tended to take the lead in attitudes toward the marketability
of the maize variety SC513. The statements “Grain of SC513 is easy to sell” and
“Planting SC513 is a waste of time and money” were rated second and third in Vanduzi;
in Machipanda the same statements were rated fourth. It is not surprising to ﬁnd farmers
in Vanduzi scoring ahead on the marketability aspects of the improved maize variety
SC513 because in Vanduzi maize output is also oriented to market, as opposed to
Machipanda, where maize output is produced mainly for home consumption.

Farmers’ attitudes toward SC513 consisted mostly of cognitive responses
expressed as beliefs by farmers about the relationship between the seed and the
production and marketability characteristics that describe the seed. It is probable that
these beliefs were developed through direct and indirect experience with improved maize
variety SC513. A direct experience might have occurred when farmers conducted and

managed demonstration plots or tried out the new seed in their own plots, and when they

120

bought the seed and sold the maize output. An indirect experience might have occurred
when farmers heard radio programs on improved maize.

Most farmers in Machipanda and Vanduzi appreciated the effect of NPK and urea
on increasing maize yield. The majority of respondents agreed that NPK and urea were
good for increasing maize yield. Farmers’ attitudes toward use of chemical fertilizer on
maize consisted of both cognitive and affective responses. The cognitive responses were
expressed as beliefs that farmers have about the relationship between the fertilizer and its
capacity to increase maize yield as well as the relationship between the fertilizer and its
worth (value) for maize production. The affective response was expressed as a feeling
(good or bad) experienced by farmers when fertilizers are paired with their effect on the
crop and soil. For example, farmers, particularly in Machipanda, expressed a negative
attitude toward the effect of NPK on the soil. These farmers might have learned the
affective response “Chemical fertilizer is not good (or is harmful) for the soil” through
direct emotional experiences that occurred when farmers stopped applying fertilizer and
then obtained low yields. As most farmers explained, when asked why they thought
chemical fertilizer, especially NPK, was not good (or was harmful) for the soil, “NPK
kills the soil. Once you use it, you should keep using it. The soil becomes dependent on
it. If you stop using NPK, your soil will not produce as well as before. . . .”

The application of NPK is viewed as an intrusion of plant nutrients into a
balanced environment. However, correct applications of commercial fertilizers have
been shown to replenish the nutrients removed by the crop and in some cases to exceed
levels of nutrients found in the soil before they were farmed (Hoeﬁ et al. 2000). Chances

are that when they stopped using NPK or used inadequate quantities of this fertilizer,

121

farmers experienced decreases in maize yield, not because NPK kills the soil but because
the nutrients removed were not replenished. Another aspect explicit in farmers’
explanations is a fear of their maize production become dependent on an input (N PK) that
farmers feel cannot afford to buy regularly.

The generally positive attitudes toward hybrid maize SC513 and the chemical
fertilizers NPK and urea may reﬂect the proﬁtability obtained with these technologies. A
study on proﬁtability of improved maize technology (J eje at al., 1998) demonstrated that
in 1996 and 1997, some farmers in the Manica province were able to make a proﬁt using
the improved technologies. Rough estimations on return to family land and labor for the
year 2006 (Chapter II) also indicated that family farms would make a proﬁt if they adopt
the hybrid maize SC513 and chemical fertilizers, provided that farmers apply the
technology as recommended, are risk takers, and prioritize the objective of making proﬁt.

Because attitudes are formed when people have information and gain knowledge
about the attitude object, the role of public extension services and private input providers
(wholesalers, retailers, and informal traders) in providing information on production
characteristics of improved maize varieties, information on marketability aspects
including price of improved maize varieties and demand for maize produce, and
information on the use and effects of chemical fertilizers should be enhanced.

This study also performed comparisons among mean attitude scores and age
categories, sources of information within each study location, and between mean attitude
scores and gender of respondents and study locations. Only in Machipanda the results
indicated an effect (p <.10) of age on attitudes toward production characteristics of

SC513. Respondents in the age category 45 to 60 years held stronger positive attitudes

122

towards the characteristics of SC513. It is possible that these respondents had longer
direct experience (through cultivation and/or processing the grain) with the maize variety
than respondents in other age categories.

The comparison among mean attitude and sources of information (neighbors,
extension services, and market) showed that in Machipanda and Vanduzi, respondents
held positive attitudes toward SC513, NPK, and urea, regardless of the information
source. Nevertheless, in Vanduzi, there was a tendency for respondents who heard about
NPK and urea fertilizers from extension services to hold stronger positive attitudes
toward their use in maize production than respondents who heard about them from
neighbors. It is not surprising to ﬁnd stronger attitudes among farmers who used
extension services as information sources because information about fertilizers is speciﬁc
in methods of application, amounts to apply, and time of application. Personal contacts
between extension agents and farmers are effective for delivering such speciﬁc
information to farmers and eventually persuade farmers to use fertilizers. These results
indicate that information on chemical fertilizers by extension services is important to
improve farmers’ attitudes toward their use. Therefore, extension services efforts related
fertilizer use should be enhanced. Findings on mean attitude by gender showed that male
respondents in Machipanda held stronger positive attitudes toward production
characteristics of improved maize variety SC513. Overall, for chemical fertilizers, male
farmers also tended to hold stronger positive attitudes toward the use NPK and urea. The
fact that extension agents and input providers were mostly men may have played a role in
male farmers strengthening their attitudes toward improved maize variety SC513 and

chemical fertilizers NPK and urea. During ﬁeldwork in Machipanda and Vanduzi, it was

123

noticed that the majority of extension ofﬁcers and input retailers were men. This fact
may have contributed to male farmers having more opportunity than female farmers to
exchange information about the technology with extension agents and input dealers. As a
result, the attitude toward technology that male farmers developed through direct
experience in using the technology is further reinforced through the interaction with
extension agents and input providers. Female farmers are left mostly with one way of
developing their attitudes - i.e., direct experience with the technology. Female farmers
are also involved with maize cultivation in the two study locations, and apparently more
in Machipanda than in Vanduzi, so it is important that women also hold strong positive
attitudes toward the use of chemical fertilizers. Females’ attitudes toward chemical
fertilizers can be strengthened through enhancement of their access to input retailers
(although this may imply a need for further analysis of gender roles to determine who has
the responsibility to buy the seed) and extension information on improved maize varieties
and chemical fertilizers. Women should be provided with fellowships to pursue higher
education in agriculture and extension education, extension services should hire women
as extension agents, and seed companies and fertilizer importers should encourage
women to work as input retailers.

When mean attitudes were compared between the two study locations, it was
found that Machipanda and Vanduzi did not differ in attitudes toward the marketability of
improved maize variety SC513, but the two locations differed in attitudes toward its
production characteristics, attitudes toward use of NPK, and attitudes toward use of urea.
Respondents in Machipanda tended to hold stronger positive attitudes toward production

characteristics of SC513. This suggests that farmers in Machipanda had more direct

124

experience in cultivating maize variety SC513 than farmers in Vanduzi. For example, for
the growing season 2005-06, the cultivation of maize variety SC513 was mostly reported
in Machipanda. The comparison between attitudes of farmers in Machipanda and
Vanduzi toward chemical fertilizers indicated that respondents in Vanduzi tended to hold
a stronger positive attitude toward NPK than respondents in Machipanda. A relatively
weaker positive attitude toward NPK indicates a tendency for farmers in Machipanda to
develop an unfavorable position toward NPK. Such a tendency might have resulted from
experiencing a decrease in maize yield when farmers discontinued the use of NPK.
Because purchased inputs, particularly fertilizers, are expensive for smallholders, it is
common to provide farmers with credit to facilitate the acquisition of inputs. For
example, successful experiences with maize intensiﬁcation, such as the smallholder-led
maize revolution in Zimbabwe (Eicher and Kupfuma, 1997), had a strong credit
component. Respondents in Machipanda and Vanduzi rarely reported having received
credit for seeds and fertilizers. Therefore, to keep farmers strongly positive about
chemical fertilizers, there should be mechanisms such as the provision of inputs on a
credit basis to help farmers regularly acquire chemical fertilizers.
Pattern oLAdqption of Improved Maize Variety SC513 and Chemical Fertilizers NPK
and Urea in the Machipanda and Vanduzi Administrative Posts

Farmers in Machipanda and Vanduzi have been using improved maize variety
SC513 since early 1990; use of chemical fertilizers for maize production started much
earlier, around 1970. The technological package consisting of improved maize variety
SC513 and chemical fertilizers NPK and urea was formally introduced in the area in

1995. Since the formal introduction of these technologies, new farmers joined each year,

125

increasing the percentage of usage (i.e., the number of farmers who have used the
technology). This increase was more rapid in the highlands of Machipanda than in the
lowlands of Vanduzi. A possible explanation is that hybrid maize seeds do better than
local varieties in high rainfall regions such as Machipanda where higher incidence of
fertilizer use is also expected than in the lowland regions (Kaliba et al., 2000).

It is important to note that the pattern of adoption provides a positive picture, in
the sense that it expresses that the number of respondents who had used the technologies
through the years was accumulating (increasing). Nevertheless, the current adOption rate
may not be high. For example, in the growing seasons 2004-05 and 2005-06, the
percentage of respondents cultivating, continually, improved maize variety SC513 and
applying chemical fertilizer on maize was lower than the cumulative percentage in both
study areas.

One obvious reason for reduced adoption is discontinuance. After they had
adopted the technologies, farmers discontinued their use for various reasons including
replacement, dissatisfaction, and misuse of the technology (Rogers, 1995). In this study,
the discontinuance of improved maize SC513 use was mainly due to dissatisfaction with
husk cover and susceptibility to storage insects and replacement of hybrid variety SC513
by OPV and local maize varieties. The discontinuance of fertilizers was due to lack of
money to purchase NPK and urea.

The increase in percentage of usage was different in the two agro-ecological
regions, and these ﬁndings call attention to the need for evaluation of the suitability of
improved maize production technology in a particular agro-ecological zone. The reasons

for discontinuance have important implications for maize breeders and social scientists.

126

Breeders need to improve the husk cover and resistance to storage insects to make Hybrid
variety SC513 more attractive to farmers, and social researchers need to investigate ways
by which farmers can be provided with fertilizers at affordable costs, including provision
of fertilizer on a credit basis.
Factors Associated with Adoption of Improved Maize Variety SC513

As expected, respondents in the highlands of Machipanda were more likely to
adopt improved maize variety SC513. Indeed, the agro-ecological zone had the greatest
effect on the adoption of improved maize variety SC513. Holding other explanatory
variables constant, it is predicted that a farmer in the highlands of Machipanda has a
predicted probability of about 73 percent of having cultivated SC513 in the period from
1995 to 2005, while a farmer in the lowlands of Vanduzi has a predicted probability of
about 9 percent (see Appendix E). These results can be related to the effect of high
annual rainfall in the highlands of Machipanda. This argument is supported by other
studies (Kaliba et al., 2000; Hintze etal., 2003). Hybrid maize may be more adaptable to
the highlands of Machipanda than to the lowlands of Vanduzi. In fact, during the ﬁeld
work, the majority of respondents in Machipanda reported planting hybrid maize in the
past growing season, but very few farmers in Vanduzi reported planting hybrid maize.
Another explanation for having high probabilities of adoption of SC513 in the highlands
may be the easy accessibility to hybrid maize SC513 in Machipanda, which is facilitated
by informal traders who sell the seed at the farmers’ doors.

Because being in the highlands increases the probability of adoption of improved
maize variety SC513, extension should prioritize educational opportunities for maize

farmers to learn about its adaptability to that agro-ecological zone. Extension should

127

investigate whether informal traders offer a cheaper and more effective means for
delivering seed to farmers.

As expected, positive attitudes toward SC513 were positively associated with
adoption of it. Positive attitudes toward improved germination and drought tolerance of
SC513 over local varieties were the second most important factor in adoption of SC513.
The results showed that a tendency to have negative attitudes toward improved
germination and drought tolerance of SC513 (i.e., when the attitude scores increase)
decreased the probability of a respondent having cultivated it. This indicates that positive
attitudes toward the production characteristics of SC513 affect its adoption positively.
For example, when a farmer scores the item with a 1, he/she has a predicted probability
of having cultivated SC513 in the period from 1995 to 2005 of about 73 percent; when
the score is a 2, the probability of having cultivated SC513 is 42 percent. Other studies
found that farmers’ perceptions of production characteristics (yield, maturity rate,
drought resistance, insect resistance, lodging resistance, grain weight) determine variety
selection and adoption (Hintze et al., 2003; Adesina and Zinnah, 1993). These results
imply that researchers should continue and strengthen research on production
characteristics of improved maize varieties, and public extension services as well as the
private sector (seed companies, wholesalers, and retailers) should emphasize messages on
the production characteristics of improved maize varieties, including seed quality,
drought tolerance, and maize meal quality.

As expected, how-to knowledge was positively associated with adoption of
SC513. Knowledge of the advantages and disadvantages of improved maize SC513 was

the third most important factor in its adoption. Holding other explanatory variables

128

constant,.it is predicted that a knowledgeable farmer has a predicted probability of having
cultivated SC513 in the period from 1995 to 2005 of about 73 percent, while a non-
knowledgeable farmer has a predicted probability of about 27 percent. Because some
non-knowledgeable farmers may be cultivating SC513, the chances are that this hybrid is
not being used properly. Other empirical studies mentioned the inﬂuence of speciﬁc
knowledge about the technology on farmers’ behavior (Lichtenberg and Zimmerman,
1999). This means that, in addition to production characteristics of improved maize
varieties, the public extension services and the private sector should also improve
provision of information on the beneﬁts and disadvantages associated with improved
maize varieties. This information should be accompanied by instructions and means to
minimize the disadvantages, such as storage insect damage.

Positive attitudes toward ease of selling grain and fresh maize of SC513 were the
fourth most important factor in adoption. The results showed that a tendency to hold
negative attitudes toward marketability of produce from SC513 (i.e., when the attitude
scores increase) decreased the probability of a respondent having cultivated it. This
indicates that positive attitudes toward the marketability of SC513 positively affect its
adoption. Hintze et al. (2003) also described a positive association between the adoption
of hybrid maize and commercialization of maize output. This association indicates that
farmers obtain some surplus from SC513 that can be used for household income. Thus,
in addition to production characteristics and beneﬁts and disadvantages associated with
improved maize varieties, extension messages should provide information on prices and
demand for it. Breeders should develop maize varieties whose output is marketable. The

private sector (seed companies, wholesalers, retailers, and informal traders) should

129

provide farmers with marketing information. For the interaction term Administrative
post * attitudes toward marketability of SC513, the coefﬁcient (B = .884) indicates that
the difference between the rate of change in Machipanda and Vanduzi increases as the
scores on attitudes toward marketability of produce ﬁ'om maize variety SC513 increase
(or as the attitudes toward marketability become negative). This ﬁnding is consistent
with the description that maize output in Vanduzi is mainly oriented toward

commercialization. It appears that, although respondents in Vanduzi lagged behind in the

 

cultivation of SC513, for these respondents commercialization of maize output is an
important source of income. In fact, farmers in Vanduzi were much more involved with .
commercialization of maize output. For example, it was common to see farmers selling I

maize directly to the only wholesaler located in the center of the village. In some remote
communities of Vanduzi, farmers had developed storage facilities where large quantities
of maize were consolidated, waiting to be transported to other places.

This suggests that if breeders develop varieties with traits that are marketable and
extension services strengthen the delivery of information on transportation, marketing,
processing, prices, and demand for maize, levels of adoption may improve, particularly in
Vanduzi. Contrary to what was expected, extension had no signiﬁcant effect on adoption
of SC513. As a source of information, extension had a positive but insigniﬁcant impact
on farmers decision to adopt SC513. It is plausible that all sources of information
(market, neighbors, extension, and market) provide farmers with useful information.

Other variables that were not signiﬁcant are age, level of formal education, and
total number of people in the family. They had no signiﬁcant inﬂuence on the odds of

adoption of SC513. Other empirical studies have found similar results (Kaliba et a1. ,

130

2000; Abebaw and Belay, 2001; Tesfaye and Alemu, 2001). The observance of no
signiﬁcant association between level of education and adoption of SC513 could be
related to the fact that improved maize varieties are simpler technologies than chemical
fertilizers. As CIMMYT (1993) argues, education is expected to be associated more with
complex technologies such as agrochemicals. Another possibility is the existence of
extension and neighbors who can provide farmers with information about SC513. For
example, Abebaw and Belay (2001) and Tesfaye, Bedassa, and Shiferaw (2001) explain
that the lack of signiﬁcant relationships between formal education and the adoption of
improved maize varieties may be related to the availability of information about the
technology and the possibility that extension information, to some extent, substitutes for
formal education in the rural areas.

Nevertheless, signiﬁcant associations between respondents’ characteristics (age
and family size) and adoption are not unusual (Abebaw and Belay, 2001; Lawal et al.,
2004; Tesfaye and Alemu, 2001; Tesfaye, etal., 2001). Thus, the mixed results suggest
that more research is needed for a better understanding of associations between
respondents’ characteristics and adoption.

Factors Associated with Adoption of NPK and Urea Fertilizers

As expected, how-to knowledge was positively associated with adoption of NPK
and urea. Knowledge about the method to apply fertilizer had the greatest effect on the
adoption of NPK and urea chemical fertilizers. Holding other explanatory variables
constant, it is predicted that a knowledgeable farmer has a 91 percent probability of
having applied NPK on maize in the period from 1995 to 2005, while a non-

knowledgeable farmer has a probability of having applied NPK to maize of 43 percent.

131

For urea, it is predicted that a knowledgeable farmer has a 96 percent probability of
having applied urea on maize, while a non-knowledgeable farmer has a 69 percent
probability of having applied urea on maize. It seems that there is considerable
probability that a non-knowledgeable farmer would apply fertilizers inappropriately,
resulting in discontinuance. The predicted high probability of adoption when a farmer is
knowledgeable about the use of fertilizer on maize is not surprising. Rogers and Havens
(1961) found that knowledge of fertilizer (i.e., how to use fertilizer and what nutrients the
crop needs) acted as an intervening variable between “attitude toward fertilizer” and “use
of fertilizer.” CIMMYT (1993) explained that chemical fertilizers are complex
technologies that require some computation skills. Because knowledge about how to use
fertilizer greatly affects the adoption of chemical fertilizer, extension services should
prioritize the improvement of farmers’ knowledge on fertilizer use. As expected,
extension was positively associated with adoption of NPK and urea. Holding other
explanatory variables constant, it is predicted that when a farmer learns about NPK from
extension, he/she has a 90 percent probability of having used NPK for maize production,
compared with a 28 percent probability when a farmer learned from neighbors. When a
farmer learns about urea from extension he/she has a 96 percent probability of having
applied urea on maize, compared to a 64 percent probability when a farmer learned from
neighbors. It seems that neighbors are somewhat effective in inﬂuencing farmers to use
fertilizers, particularly urea. Other studies had found positive relationships between
extension and use of chemical fertilizers (Abebaw and Belay, 2001; Tesfaye et al., 2001;
Kaliba et al., 2000). This indicates that strengthening the provision of information on the

use of fertilizers through extension (i.e., results from demonstration plots, and individual

132

or group meetings) may improve the levels of adoption of chemical fertilizers in
Machipanda and Vanduzi. Moreover, given the relevant role played by extension in the
adoption of chemical fertilizers, one option would be to concentrate extension resources
to train farmers in the optimal use of fertilizers. Another option would be to promote
vocational training on improved maize varieties technology, including chemical
fertilizers, in the rural schools.

As expected, respondents in the highlands of Machipanda were more likely to
adopt NPK and urea than respondents in the lowlands of Vanduzi. Holding other
explanatory variables constant, it is predicted that a farmer in the highlands of
Machipanda has a 84.9 percent probability of having used NPK for maize production and
a 92.6 percent probability of having applied urea on maize; a farmer in the lowlands of
Vanduzi has a 70 percent probability of having used NPK for maize production and a
74.4 percent probability of having applied urea on maize. Other studies have also
hypotheszed a higher incidence of fertilizer use in regions where rainfall is relatively
higher than in the lowlands (Kaliba et al., 2000). In Mozambique, the highlands are
prone to erosion and loss of soil fertility due to high levels of precipitation (PROAGRI,
2003). Hence, farmers located in the highlands (Machipanda) would be more likely to
adopt NPK and urea than farmers in the lowlands (Vanduzi). Likewise the use of NPK
and urea for maize production appears to be more acceptable in the highlands of
Machipanda than in the lowlands of Vanduzi. Because of the visible effects of erosion
and loss of soil fertility in the highlands, farmers in Machipanda may be more aware of
fertilizer application than farmers in Vanduzi. Therefore, extension efforts related to

fertilizer use should be intensiﬁed among farmers in Vanduzi. Respondents’

133

characteristics (age, level of formal education, and total number of people in the family)
and farmers’ attitude toward NPK and urea have no signiﬁcant inﬂuence on the logarithm
of odds of adoption of NPK and urea. Other empirical studies had found similar results
regarding respondents’ characteristics. Tesfaye and Alemu (2001) found no association
between formal education and adoption of chemical fertilizers. It seems that even for
fertilizers, which are considered a more complex technology (CIMMYT, 1993) than
maize varieties formal education had no signiﬁcant inﬂuence on their adoption. As stated
before, the availability of information about the technology and the extension resources
to some extent substitute for formal education in the rural areas and may explain the lack
of relationship between formal education and adoption of chemical fertilizers (Abebaw

and Belay, 2001; Tesfaye, Bedassa, and Shiferaw, 2001).

134

CHAPTER VI

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

This chapter is organized in ﬁve sections. The ﬁrst section summarizes the
rationale for the study, the purpose of the study, and the procedure used to collect,
examine and analyze data. The second section presents the main conclusions of the
study. The third section draws practical implications for dissemination of improved seed
and chemical fertilizers in the study area. The fourth section presents suggestions for

further research. The ﬁfth section presents agricultural devepment policy-related issues.

Summary

In Mozambique, political leadership and agricultural development policy favor
adoption of improved maize varieties and chemical fertilizers to increase productivity of
maize. One of the challenges to agricultural development policy is to promote
widespread adoption of improved maize varieties and chemical fertilizers. Despite low
maize yields and availability of improved maize varieties and chemical fertilizers to
improve maize production, these technologies are still not widely adopted in
Mozambique. This leads to questioning why farmers are apparently unwilling to adopt
improved maize varieties and chemical fertilizers. Research that addresses this question
is still limited in Mozambique.

The overall purpose of this study was to determine the inﬂuence of farmers’

characteristics and farmers’ attitudes toward improved maize and chemical fertilizers on

135

adoption of improved maize SC513 and NPK and urea fertilizers by farmers in the
highlands of Machipanda and the lowlands of Vanduzi administrative posts of the Manica
district. This study involved a cross-sectional survey. A questionnaire, consisting of 26
items, was developed and reviewed for content validity by a panel of experts, which
comprised academic advisory committee members and an expert on survey research from
the Ministry of Agriculture in Mozambique. After revisions were made, the survey was

pilot tested on 22 heads of households from Machipanda and Vanduzi. This allowed for

 

evaluating the appropriateness of the questions given. The questionnaire was I“
implemented through personal interviews with 120 randomly selected heads of
households from Machipanda and 173 randomly selected heads of households from I

Vanduzi. The interviews were conducted during April and May 2006. Data were
examined and analyzed using SPSS 15 and AMOS 7.

The statistical methods used in this study included:

Reliability, Exploratory Factor Analysis and Conﬁrmatory Factor Analysis.
These methods were used to construct summated scales and test their reliability.

Descriptive statistics. These methods were used to describe the characteristics of
respondents and patterns of adoption of improved maize varieties and chemical
fertilizers, and to assess farmers’ attitudes toward improved maize varieties and chemical
fertilizers.

T-test and one-way ANOVA. These tests were used to compare mean attitude
scores between the two study locations and among age categories, sources of information,

and respondents’ gender within each study location.

136

Multiple Logistic Regression. This technique was used to determine the factors in
adoption of improved maize varieties and chemical fertilizers. Eight hypotheses were
formulated to examine the factors associated with adoption of improved maize varieties
and chemical fertilizers. The effect sizes were used to interpret the relative importance of

factors associated with adoption of improved maize varieties and chemical fertilizers.

Conclusions

The main conclusions of this study are as follows:

The majority of respondents in the highlands of Machipanda were female; the
majority of respondents in the lowlands of Vanduzi were male. The majority of
respondents in both study locations were 44 years of age and younger. In general, the
interviewed farmers were heads of relatively large families, with an average of eight
members per family. The level of education among the respondents was low. Most had
some primary school but a considerable number were illiterate, particularly in the
lowlands of Vanduzi. To learn about the improved maize variety SC513 and chemical
fertilizers NPK and urea, respondents used mainly neighbors and extension services,
including radio broadcasting on maize variety. In both study areas, most respondents
were knowledgeable of the advantages and disadvantages of improved maize varieties,
but few considered themselves knowledgeable about application methods of NPK and
urea fertilizers on maize, especially in the lowlands of Vanduzi.

Respondents tended to hold a generally positive attitude toward the marketability
of improved maize variety SC513, production characteristics of SC513, and use of NPK
and urea for maize production. Farmers’ positive attitudes may be explained by

proﬁtability of the improved technologies. The attitude strength varied according to the

137

 

 

location (Machipanda and Vanduzi) and within each location according to gender and
sources of information (neighbors, extension, and market). Male respondents in
Machipanda tended to hold stronger positive attitudes toward production characteristics
of improved maize variety SC513 and use of urea for maize production. Male
respondents in Vanduzi tended to hold stronger positive attitudes only toward the use of
chemical fertilizers NPK and urea for maize production. The sources of information did
not have an effect on attitudes of farmers in Machipanda. In Vanduzi, farmers who
learned about chemical fertilizer from extension services tended to hold stronger positive
attitudes toward chemical fertilizer than farmers who learned about this technology from
neighbors.

The results showed that the number of farmers adopting improved maize variety
SC513 is higher than the number of farmers who ever used NPK and urea for maize
production. After these technologies have been adopted, discontinuance occurred. Few
respondents, particularly in the lowlands of Vanduzi, planted improved maize SC513 and
applied NPK and urea on maize during the last growing season, 2005-06. The major
reasons for discontinuance of Hybrid maize SC513 were non-complete closure of the
husk cover, which exposed the cob to rain and attack by ﬁeld insects and diseases;
susceptibility of grain to attack by storage weevils, including Protesphanus truncatus and
Sitophilus zeamais; lack of money to purchase seed of SC513; non-availability of seed of
SC513; and preference for other improved varieties such as SC700, Matuba, and local
maize varieties considered better than SC513. Lack of money to purchase fertilizers was
the most reported reason for discontinuing NPK and urea. The inappropriate use of

technologies might have also been a reason for discontinuance. There were considerable

138

probabilities of adoption of the improved maize variety and chemical fertilizer by farmers
who were not knowledgeable of the advantages and disadvantages of improved maize
and methods of fertilizer application.

Farmers’ characteristics and farmers’ attitudes toward the marketability of
improved maize variety SC513, the production characteristics of SC513, and use of NPK
and urea for maize production inﬂuenced adoption of these technologies in Machipanda
and Vanduzi. The agro-ecological zone and how-to knowledge were common important
factors associated with the ad0ption of improved maize SC513 and NPK and urea
fertilizers. Nevertheless, agro-ecological zone was the most important factor in the
adoption of improved maize variety SC513, and how-to knowledge was a decisive factor
in the adoption of NPK and urea fertilizers. The three most important factors associated
with adoption of improved maize SC513 were: agro-ecological region, attitude toward
the production traits of improved maize SC513, and knowledge about advantages and
disadvantages of improved maize varieties. The three most important factors associated
with adoption of chemical fertilizer NPK were: knowledge about the method to apply
fertilizer, extension on NPK, and agro-ecological zone. And the three most important
factors associated with adoption of chemical fertilizer urea were: knowledge about the
method to apply fertilizer, agro-ecological zone, and extension on urea.

This study showed that farmers in Machipanda and Vanduzi do support the use of
improved maize varieties and chemical fertilizers. Nonetheless, climate and soils (agro-
ecological zone) and the knowledge of advantages and disadvantages of improved maize
varieties as well as knowledge to use chemical fertilizers had greater inﬂuence on the

decision to adopt them. Respondents’ characteristics, such as age, level of formal

139

education, and total number of people in the family, did not inﬂuence signiﬁcantly the
adoption of improved maize variety SC513 and chemical fertilizers NPK and urea in

Machinpanda and Vanduzi administrative posts.

Implications

The results of this study have implications for the National Directorate of Rural
Extension (DNER), the private sector, and national and international research institutions
engaged with dissemination of improved maize varieties and chemical fertilizers in the
highlands of Machipanda and lowlands of Vanduzi. Other institutions for which the
results of this study have implications are the institutions of formal education and non-
governmental organizations (N GOs) engaged with non-formal education.

Improvement of How-to Knowledge_on Fertilizer Application and Knowledge of
Characteristics of Improved Maize Varieties

How-to knowledge is an important determinant of adoption. When farmers do
not know how to use a technology, the odds of adoption are very low. This study showed
that a considerable number of farmers felt they were not knowledgeable about methods of
applying fertilizer on maize, and some farmers were not knowledgeable of the
advantages, and disadvantages of improved maize traits. Nonetheless, the predicted odds
of non-knowledgeable farmers having cultivated Hybrid maize SC513 and having applied
NPK and urea on maize were not very low, so some farmers may have been using the
technologies inappropriately. Therefore, the DNER should collaborate with specialists in
soil fertility management to design training programs on optimal use of chemical
fertilizers. It may also be important for the DNER to consider improving farmers’

knowledge of the advantages and disadvantages of characteristics of improved maize

140

varieties. To improve how-to knowledge, the extension services should use methods and
results demonstration plots at the farmers’ ﬁelds.
Prioritization of Production and Marketabilitv Characteristics of Improved Maize
Varieties

Attitude assessment offers a valuable procedure to generate measurements of
support for agricultural technology. This investigation showed that farmers in the
highlands of Machipanda and the lowlands of Vanduzi were supportive of improved seed
whith enhanced germination and drought tolerance that provides a quality maize meal.
Moreover, farmers were positive toward hybrid maize SC513 because they believed that
the output from SC513 could be easily sold. This expresses support of improved maize
varieties whose output is marketable. Given these results, the following implications
were drawn for breeders, public extension services, and the private sector:

(i) Maize breeders should prioritize production and marketability aspects
of the improved maize varieties (Hybrids and OPVs), including
enhanced germination, drought tolerance, and maize meal quality. In
addition, breeders should develop instructions on seed handling to
minimize deterioration of seed quality.

(ii) Public extension services should stress information on seed quality and
output price and demand, especially in radio broadcasts.

(iii) The private sector (seed companies, wholesalers, and retailers) should
also emphasize production characteristics and marketability of
improved maize varieties when advertising and selling improved maize

varieties.

141

 

 

Provision of Fertilizers and Seed on aL Credit Basis

The adoption pattern showed that chemical fertilizers NPK and urea were less
widely adopted than the Hybrid maize SC513 and in the last growing seasons (2004-05
and 2005-06), very few respondents applied chemical fertilizer on maize. The
discontinuance of use of chemical fertilizers was mainly due to lack of purchasing power.
The DNER and social scientists need to investigate ways by which farmers can be
provided with inputs, particularly fertilizers, at affordable costs, including provision of
fertilizer on a credit basis.

Strengthening Attitudes of Female Farmers

Women are engaged with maize cultivation, particularly in Machipanda.
However, women’s positive attitudes toward SC513 and chemical fertilizers were
somewhat weaker than male attitudes. Stronger positive attitudes have the advantage of
being difﬁcult to change. Personal communication between female farmers and female
extension agents and input providers is an option to strengthen female farmers’ attitudes.
Therefore, higher agricultural education institutions should promote women’s education.
The DNER should employ women extension agents. The private sector, such as seed
companies, should promote women wholesalers and retailers.

Improvement of Traits of Improved Maize Varieties and Adaptability of
Improved Maize and Chemical Fertilizer

The results of this study indicated that farmers discontinued the cultivation of
SC513 because of dissatisfaction with its husk cover and susceptibility to storage insects.

Therefore, maize breeders should improve the husk cover and resistance to storage

142

insects to make the Hybrid variety SC513 more attractive to farmers. The DNER should
collaborate with entomologists to develop messages on how to control storage insects.

The agro-ecological region was an important factor associated with the adoption
of improved maize variety SC513 and NPK and urea fertilizers in both study areas.
Agricultural research institutions (such as Instituto Nacional de Investigacao Agraria
[IIAM] and CIMMYT) should prioritize and strengthen the development of improved
maize varieties and recommendations for fertilizer application which are attuned to agro-
ecological region (rainfall and soils). Moreover, the research could also ﬁnd varieties
that have wide adaptation, growing well in both areas.

Strengthening Coordination Amongthe [gricultural Actors

Efforts should be made by extension, the private sector, and research institutions
to establish an active web of agricultural information and knowledge for coordinating
knowledge-led agricultural development through production and exchange of information
and improvement of farmers’ knowledge on how to use the technologies.

This active web of agricultural information and knowledge would comprise the
relationships among farmers and providers of agricultural services, including agricultural
extension, marketing analysts, private input providers, seed companies, and maize
processors (maize millers, and the animal feed industry). The term “active” indicates a
web in which personal communication is preferred and used over other means of
communication. “Information” stands for the elaboration of persuasive messages and
communicating them to the users, and “knowledge” refers to the educational component
of the web, which ensures that farmers have adequate how-to knowledge to use the

technologies.

143

Suggestions for F urther Research

As previously mentioned, agricultural development policy in Mozambique favors
adoption of improved maize varieties and chemical fertilizers to increase maize
productivity. Therefore, knowledge of factors associated with the adoption of improved
maize varieties and chemical fertilizers is vital to improve agricultural development
policy. The following studies are suggested:

Study 1. Effect 0[ Market Inﬁastructure and Role of Local Institutions on
Ad0ption of Improved Maize Technologies

The current study investigated a few factors associated with adoption of improved
maize varieties and chemical fertilizers in two speciﬁc locations, the highlands of
Machipanda and the lowlands of Vanduzi. To provide information that reﬂects the entire
country, it may be important to scale up the study to cover more regions and determine
the effect of additional factors, including access to market infrastructure (roads, transport,
and marketplaces), access to credit, family size (labor), farm size, level of formal
education, and the role of local NGOs and farmers’ groups in coordinating and
integrating production and marketing activities, as well as local resource mobilization.

Study 2. The Economics of Improved Maize Technologies

This study did not address the economic proﬁtability of Hybrid maize SC513, and
chemical fertilizer technology, which is an important factor in adoption. Therefore, a
study on economic proﬁtability of improved maize technologies is recommended. This
study can be conducted as a replication of the proﬁtability study conducted by Jeje et al.

(1998)

144

StudLi. Improvement of How-to Knowledge on Improved Maize Technologies

The results of this study showed that how-to knowledge was an important factor
for adoption of improved maize SC513 and chemical fertilizers NPK and urea. A
considerable number of farmers did not consider themselves knowledgeable about the
advantages of improved maize traits and application methods for chemical fertilizers.
This situation is a threat to adoption in two ways: ﬁrst, non-knowledgeable farmers
would tend to reject the technologies; second, the non-knowledgeable who try using the
technologies may use them inappropriately and soon become dissatisﬁed and discontinue
the technologies.

Making sure that farmers have adequate levels of how-to knowledge is very
important. Therefore, it may also be important to conduct a study to examine how
farmers in Machipanda and Vanduzi can improve their knowledge on characteristics of
improved maize and application of chemical fertilizer, and evaluate the extent to which
strengthening the crop breeding and soil management curriculum in vocational education

could be used to build knowledge on variety traits and proper fertilizer use.

Policy-related Issues

“How Can the Poor Be Reached? ”

The researcher would like to begin this section with some statements made by
William Easterly in his recent book, “The White Man’s Burden”, published in 2006. The
researcher found Easterly’s statements very enlightening for aid-dependent countries
such as Mozambique. Easterly starts by identifying two tragedies of global poverty as

follows:

145

“UNITED KINGDOM CHANCELLOR of the Exchequer Gordon
Brown is eloquent about one of the two tragedies of the world’s poor
He called for a doubling of foreign aid, a Marshall Plan for the
world’s poor, and an International Financing Facility (IF F) against
which tens of billions more dollars toward future aid could be
borrowed to rescue the poor today. He offered hope by pointing out
how easy is to do good. Medicine that would prevent half of all
malaria deaths costs only twelve cents a dose. A bed net to prevent a
child from getting malaria costs only four dollars. Preventing ﬁve
million deaths over the next ten years would cost just three dollars for
each new mother. An aid program to give cash to families who put
their children in school, would cost little” (pp. 3-4).

This is the ﬁrst tragedy of the global poverty —- that easy, low-cost solutions are
available to help the poor (Easterly, 2006). His message is that it would be easy to ﬁght
against poverty if the solutions were implemented as planned. Regrettably, what has
been planned hardly ever gets implemented, and the good intentions of targeting the poor
remain unfulﬁlled. In the description of the second tragedy of global poverty, Easterly
expresses his concern with failure to target the poor:

“Gordon Brown was silent about the other tragedy of the world’s poor.
This is the tragedy in which the West spent $2.3 trillion on foreign aid
over the last ﬁve decades and still had not managed to get twelve-cent
medicines to children to prevent half of all malaria deaths. The West
spent $2.3 trillion and still had not managed to get four-dollar bed nets
to poor families. The West spent $2.3 trillion and still had not
managed to get three dollars to each new mother to prevent ﬁve
million deaths It’s a tragedy that so much well-meaning
compassion did not bring these results for needy people” (pp. 4).

The poor can be helped through aid. However, experience has shown that it is
difﬁcult to channel the aid to the poor. As Easterly stated it, “. . .I and many other like-
minded people keep trying, not to abandon aid to the poor, but to make sure it reaches

them” (p5). One of the difﬁculties in getting aid to the poor, as Easterly pointed out is

the fact that aid for the poor is commonly channeled through big plans. Big plans

146

 

distract planners from concentrating on speciﬁc tasks that address local problems. For
this reason, Easterly suggests, people working to help the poor should have the
mentality of searchers who try to ﬁnd answers to the question “What can foreign aid do
for the poor?” rather than “What does the end of poverty require of foreign aid?”
Searchers adapt to local conditions, ﬁnd what reality is at the bottom, and hope to ﬁnd
answers to individual problems only by trial and error experimentation. Feedback fom
the poor and accountability of searchers to the poor are key elements for the success of
the searchers’ work (Easterly, 2006).

Now, how do Easterly’s concerns relate to the current study? Easterly’s ideas
relate to the current study because it is about poor maize growers in a poor country where
efforts are being made to ﬁght rural poverty through dissemination of improved maize
technology to increase maize productivity. The Agriculture Development National
Program (PROAGRI), a US $200 million phase of a long-term program (World Bank,
2004), is an example of donors’ funded program to improve the lives of million of rural
poor. There is a lot to learn from Easterly’s ideas on how to experiment with targeting
the poor.

Access to credit by the farmers is an area where Easterly’s ideas could be
experimented. For example, this study found that farmers who had adopted Hybrid
maize SC513 and chemical fertilizers NPK and urea for one or more years tended to
discontinue the use of these technologies because they lacked money to purchase them.
Successful experiences with adoption of hybrid maize and agrochemicals had a strong
credit component for acquisition of inputs. An example is the case of the successful

smallholder-led maize revolution in Zimbabwe (Eicher and Kupfuma, 1997). In

147

Mozambique, however, there is limited knowledge of best practices on production credit
(World Bank, 2004). Good repayment rates (about 98 percent) have been reported for
credit suppliers such as the V&M Grain Company, a leading domestic agribusiness
company in Mozambique, which supplies a sizable portion of farmers’ credit needs and
promotes linkage among producers associations, processors, and exporters (World Bank,
2004). An example of a relatively simple credit scheme which operated with individual
farmers at the local level is the network of input providers used by the DNER/SG2000
dissemination program. Participating farmers managed credit plots of about 0.5 ha.
Farmers were delivered inputs on a credit basis, involving nearby input suppliers and
rural traders. At the end of the cropping season, farmers had to pay back the loan with
part of their production. The credit scheme did not always work well, however. Some
individual farmers failed to pay back the loans. During the ﬁeldwork, the researcher had
the opportunity to talk to one of these farmers and ask him why he failed to pay back the
loan. His answer was that his maize crop was burned by the sun, meaning that he had a
crop failure. Some of the reasons for this could be insufﬁcient rainfall, late planting
because the inputs were supplied late, and/or misuse of chemical fertilizers.
Interestingly, this same farmer kept asking the researcher “to be a searcher” and look for
other projects that could supply him with credit. This act expressed the farmer’s
commitment to reinvest in agriculture and his interest to discuss and learn about what
went wrong with the previous credit experience.

To summarize this section, the researcher returned to the question “How can the

poor be reached?” The experiences of the V&M Grain Company and the DNER/SG2000

148

extension program suggest the following ways the poor can be supplied with credit for
inputs:
Make resources (aid) available at the local level for small maize-related proiects to
address local problems

The resources (cash money, inputs) to be supplied on a credit basis should be
available at the local level - for this speciﬁc case at the administrative post and village
levels. Seed companies should encourage local seed representatives, wholesalers,
retailers, and informal traders to increase their efforts to reach the farmers. Fertilizer
importers should channel fertilizers to the local retailers. Private banks should have
representatives at the local level.
Provide affordable interest rates and timely credit

Farmers should be supplied loans with low interest rates to promote the use of
improved maize and chemical fertilizers. Moreover, the loans should be supplied before
the growing season.
Strengthen knowledge and experience in growing maize

The seed representatives, wholesalers, retailers, and informal traders should be
knowledgeable of and advertise those aspects that make the improved seed and fertilizers
attractive to farmers, such as marketability and production traits of the improved maize
seed, and the increased maize yield obtained with use of fertilizers.
Encourage producer organizations and institutions (rules)

Because it may be difﬁcult to collect loans from individual farmers, some kind of
peer pressure is needed among the farmers to encourage loan repayment. Helping

farmers organize in producer associations or cooperatives may improve farmers’

149

 

accountability to the credit suppliers. It is the researcher’s opinion that the “searchers”
and the farmers should be accountable to each other. Farmers receiving credit should be
responsible for the possible output they will obtain with the investment. And the credit
suppliers and farmers should agree on how to share the risks associated with agriculture
production, particularly environmental risks such as erratic rainfall. Besides providing
peer pressure, producer associations are advantageous in addressing constraints such as
transaction costs and unacceptable levels of lending risk for processors and other buyers
dealing with small farmers (World Bank, 2004).
Use exteasion services to educate and get feedback from farmers

Extension agents have to get feedback from farmers and conduct regular
inspections to make sure that farmers who are supplied with credit are complying with
the basic crop production operations such as row planting, correct plant density, correct
fertilizer rates, and timely weed control to attain high yield.
Building capacity among extension agents

Extension agents should be provided with training on the basic elements of cost-
beneﬁt analysis, project design, implementation and evaluation, and marketing to
strengthen their capacity in these areas so they are able to assist farmers in a more
comprehensive manner. Current assistance is geared more toward crop production
operations.

In summary, the policies of government, the private sector, and entities supporting
poor maize growers should be carefully developed, instituted, and evaluated to address
Easterly’s concerns for getting resources (ﬁnancial and technical information) directly to

the poor.

150

 

APPENDICES

151

Appendix A: Map of Mozambique

 

152

Mozambique

 

Source: http://images.google.com
Figure 1. Mozambique and its bordering countries.

153

Appendix B: Maize Technology Package, Central and Northern Mozambique

154

 

Technology Package, Central and Northern Mozambique

Maize (grain) Target Yield: 4 ton/ha

 

1. Land Preparation: Zero Tillage
- Apply 4 liters of RoundUp per hectare 15 days before planting.

2. Maize Varieties:
- Sussuma, Manica, SC 513.

3. Seed T reatement:
- Use Regent 0.03% at rate of 40kg/ha at planting to control termites.

4. Seed rate and spacing:
- 50,000 plants/ha (25-30 kg seed/ha).
- Spacing 0.8 m x 0.25 m, one seed per hill; or 0.8 m x 50, two seeds per hill.

5. Weeding:
Apply Bullet mixed with glyphosate at preplanting and a corrective weeding when
justiﬁed.

6. Fertilizer:
- Apply a basal fertilizer at a rate of 100 kg/ha of NPK (12-24-12) at planting and top
dress of urea at 200 kg/ha 30 and 60 days after emergence.

7. Control of stem borer:
- Apply 0.3 lit/ha of cyperrnethrin in all regions. Do two treatments - at vegetative and
ﬂowering satges.

8. Irrigation:
- First irrigation immediately after sowing to guarantee good germination.
- Four irrigations could be enough every three weeks.

9. Harvesting:

— Maize should be harvested when the grains attain physiological maturity. At this stage,
maize should be treated with Actellic against weevils.

155

Appendix C: Letter of Consent

 

£7.315 .

156

ADOPTION OF IMPROVED MAIZE VARIETY AND CHEMICAL FERTILIZER BY
FARMERS OF MACHIPANDA AND VANDUZI ADMINISTRATIVE POSTS IN
MOZAMBIQUE

February 2006

Dear Farmers: (<5-(ligit code number>)

I am a faculty member at the Eduardo Mondlane University in Maputo. I am pursuing

graduate studies at Michigan State University, USA. As part of my graduate studies, we

are conducting a study about the adoption of improved maize production practices by

farmers in Manica District Of Mozambique. The purpose of this study is to ﬁnd out the

level of adoption of improved maize production practices by the farmers so we could !
make recommendations to improve the extension service in this area. , ...

We are conducting this survey with farmers in Machipanda and Vanduzi Administrative
Posts. Your name was selected on a random sample. We would like to ask you a few
questions about maize production practices you follow on your farm. It will take us _. ,.
about 30-50 minutes to complete this interview. Please note that there are no right or F!
wrong answers, we just need your Opinion. All your answers are completely voluntary

and will remain conﬁdential. They will be analyzed together with those of other farmers.

Only the combined results will be used in this study. You indicate your voluntary

agreement to participate in the interview.

 

We will prepare a research report at the end of the study. If you like to receive a copy of
the research summary, please let me know and we will arrange to send you a copy.

If you have any questions about this study, please contact Dr. Murari Suvedi at Michigan
State University. If you have questions or concerns regarding your rights as a study
participant, or are dissatisﬁed at any time with any aspect of this study, you may contact
Dr. Peter Vasilenko., Chair of the University Committee on Research Involving Human
Subjects (UCRIHS).

Dr. Murari Suvedi Dr. Peter Vasilenko

Michigan State University Michigan State University
Department of CARRS UCRIHS

135 Natural Resources Building 202 Olds Hall

East Lansing, MI 48824-1222 East Lansing, MI 48824

PH: 517-432-0265 PH: 517-355-2180 FX: 517-432-4503
suvedi@msu.edu irbchair@ores.msu.edu

 

 

Thank you for your time and cooperation.

Sincerely,
Murari Suvedi, Professor
Eunice Cavane, Graduate Student

157

 

Appendix D: Survey Instrument

 

158

 

ADOPTION OF IMPROVED MAIZE VARIETIES AND CHEMICAL FERTILIZER IN
MACHIPANDA AND VANDUZI ADMINISTRASTIVE POSTS OF THE MANICA
DISTRICT, MOZAMBIQUE

(INTERVIEW SCHEDULE)

 

This survey is aimed at collecting information on adoption of improved maize
varieties and chemical fertilizers. THE INFORMATION PROVIDED WILL BE
USED ONLY FOR RESEARCH and for planning educational programs on

 

maize production.

 

 

 

Researcher: Ms. Eunice Cavane, Doctoral Candidate
Department of Community Agriculture and Resources Studies
131 Natural Resources Building
Michigan State University
East Lansing, MI 48824-1222

159

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ZO—H<U—EHZHG~ Elma—20mm:

.oﬁtqeqqm a we: :23 EBB Em .EoEquam cm Em m_nm__m>m 8: m. 8&8 85:26 9: E

Egmgoﬁmq >m<Fz 30> Em 38:8 no Logo. 9: 58m 288qu m5 E95 235% 2 8&8 BEES 9: a

Aconcagua 336 SF sauna—mam

comm—Eon use SEQ 3 poem .3 090 05 Bone 320% .385 823:3 23 .coﬁoa 05 .o. a cocoseoa 368 53> 3302:
608 mm 0:3 ”8&3 05 5F? :ocwmcgcoo a 3m: 9 8:— 283 F ONES wEBBw Bantam 8m £8526 >023 26m a

022.: Dioramas DEN—wcoE 8.335 2: Low x83 F new mm 082. F2 3:626 .. :ooEotﬂ wEEoE BOO

 

”828mm: 9 8:28. m. 4,3920; 9.: n: szmmmo Em 328 see on :3 5g 832: Sizes geoa
$62.8 $35 o5 282 2 >m<wwmowz m3: ME 92:

85325: 2m :2 we: :08 2:550 228255 2: 2m 20:83 m5 9mm:

mZO—HUDy—Hml— mmamgmm—HZF EHZHU

o n v n N F e a o F m r. v n N F e F
N N N N N N N F F F F F F F F F F a a N o n v n N F s.
a. .m _ . _ .._. s _ a ._ o > o. . ...,. to o a.” ...z _.. e .s, -_.: a ._ ....
, o m ._ < o g _ z a a m a a __ u < a a a m < a a o a a
or EN: ..u -o m b... . < :0 ...z: ..s .a: .5 Or .2 Sun 5: 2E -5 c, -Nr N. -5 in. o o s 8.. .5.
8.5.5
..oEEoFE 6:5 .98.. .8 «:32

ill

160

 

 

 

 

 

 

 

 

 

 

 

 

 

A3 «Esnmam
AG mum Um
5 mm 8:32
a. a
Snag § 225
E :22on
owe 28m
A 23323230 agmﬁm
E F88 88 av 28% antes
E 5% Mowmﬁuosmo E 385902
m 32 $50 .
E 98% ages E E5... .83
E 2 E _ 3.8 mmeza ﬂ :3
_ 068 yum—Elm: MN 730 $2.282: so» mowotg 3 cu 8%
23:22.2: so» 832:; 3 cm a?» =— MN BEE 05 Sosa Bo: :— MN
838 2: 8m .5» E 223 .«3 2: oz 5 we, 8» Eu 823 so; .«3 8V oz 3 m;
.822... 2: .232... 2.. 3%: ﬂags»? Bee

 

amm— dN £331? 338 wEBo=£ 05 =on .30 20% 263 .m0

 

 

 

wEBo=£ 2: Sons v.80: 55 so» 0%: .NO

 

Ta?» ”FHHZH— .MN momma... :30 59A @5323 .83 so» Eu 555 .—O
.wcuawa

:89 32 no» 83253 SEE F0 090 05 was “388 955323 E oocotoaxo Box Sons aso» Eat Emo— 8 8:— 283 _ JEE

war—H535 ENE 98:33:: m0 ZO—EOG< G72 HQGHABOZM “F ZO—HUm—m

161

 

 

 

 

 

3:0ng

as $50

a 5 mm «35: .ueaumam
s: 2:33 .38

as am ~25: 558

a: £338 $2555
:5 2... um 82:35
3: mm 822,. 825.55

 

 

Q 5523 am on m2“.

. av 89 B 25

3383 E .26 E 5N m

guano .2 855% s 8255 Ag 5» ow
§ 985 .050 s 8.08 E 22%

E egg ﬁsh—E E2. 88:85 .3 «E325
a $.53 550% 22,. 8865; 6 2m om
g 5%,: we 29. 886.5 5 mm 82»:

3 ~88 .88 2.. so: 8355 E 90:58
§ .822 as 52 85 § 825220

_ 2.8 mmeza ﬂ _ $8 amaze. ﬂ
25.53 :0» being of mo 33 05 2368 so» Ev 20:3 .30 but? 332 .v0

 

72¢ £5 6.88.. .8392;

05 no 05:: 2: 52—8059. we: 33. .8855 :— .n~ mmoom .80 I BO .833. wEBEw «mg 05 :83 so» Ev mozotg 365 «as? .v 0

162

 

SO - e0 :3 «on a £0 - 50 in £53.25 .5 2m Um #5 3:32 @8529 hth«._ 5233. MERE» «ma. 2: E a: .nw

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

v
m
m
_
€255.28
gigs 3 ”MM“
5 as. E 82% |Saamv E x :52 755%
E n: E 855 E 926 .E 53 __< .
BEE ﬂ $.28: 9:5: ﬂ cm 8» E ﬂ b3? :55.
E 8; 25: E 8> . M»
“552: EoEaSco voﬁwt: av 02 SEE 355::
£30m o «:3 0 “cow: 5 o c055 O.— OH DEN m . .5 Es:
w w % a. c . a a u _ 9:? a
.umo 3:30:30 ...3 $55 .80 ...mo 8%35 .93 coanoi .nmo as: .53 8E

 

 

El omO bu: in £83.25 .8 2m Um .mm «2.32 :3:— 8: Eu hon—ham £8an wEBEu .2: on: B u: MN

go:

«HERE 9 3:323 203 32m ~32: Bo: 598% 9:38» “mm. 2: 5 .w.a m .0

163

 

 

A3 «533m

 

5 2m Um

 

5 mm 8:52

 

we we mo No 5 oo oo mo no co

 

 

 

 

 

 

 

 

 

 

 

CV mu> 8V OZ :28» 03:: wESv 332:; 3:2: @399: ES: :0» ED :50

 

but? 38E 335::

mEEmE vutﬂm 30> .30

 

822:? 3:2:
:3an—

 

 

F3553

3:2: 335:: 3:52: :25on :o @355 so» 85m man: who» >82: So: in: nu?» .233 non—82:0.— «o: 83. hour—am 5. MN

7.33 05 E Augm— MN $55ng :0» 3:01? 0:2: 353:: wEEaE Sam :9» E: :25? 5-: b .0

 

A3 «Eammsm

 

5 an 8

 

5 mm «252

 

.30 3 an 2. E ﬂ

«832:3 3:2: BREE: BEE: :30: b5) .30 SV 02

.3 - 5
3 on no» a: MN

 

38>

 

 

 

SEE wEBo=£ 05 BEE: .55 so» oz“: u 0

7.5:? 59: ﬂ

£251.?

 

 

 

 

 

4
6
l

 

855830 2 m Um

 

 

 

 

 

20.850 Mm «252
Sisszoll mm SEEI
A. €85 a c .25
2 a: B3 muse“.
A V on 2 3:590:88
. 2 $82 a .2... 9 ﬁgs. 2 22%”.
” 368$ c c 850 av 9.5:: 88
a: 83 ousoa .2 ”80 g 08.58 88
av 2903 26.. E a: 9: 9 =9, 8% 9.8:
A8 .8 :95 8V .8 9m
E 3 ms 2 =2, 3% .2. 8 9.8: A3 8552.8 5.9 .0 38.305 88
§ «.625 mama: E xomsm 2 85:8. 33 g cage;
E 8:9Eotoq 55> .o 23.265 v.03 8v 8:822 2990
E Ema =me g €29: 28m
5 8:222 2956 :3 E 8:33 88
av £52 23
E 88:85 33 729::

T258 2: .3 no>_w 2.3:? 2: 1.32. .MN 2: 3 :03» 23:3 2: :32— MN

 

 

abuts, abet? @2535

88%.. 23c $55 23 2: 2a :23 .a .m .0 2:0 $55 Bow 2: 2m ES» .3 .0

 

 

 

 

_mo:m_..29a.§:o 39:9» 2: 22:0 can .332;

032: 3.9.0.3.:— 2: 4:2: 322. 033 2: a: .MN momoozo 30% 2:03 28 £023 3:033 052: 5233 88:0 8 was so» .3 .w .0

 

 

~23» 22: E mo:o_.8> 8:2: 335:: @553 no: 29A 2: >53 .05 .0
”78:85.. 2533:: 2: an: .25 8:22.; 032: BEE—Em 22:5“... 6253» 38...: Eu?» 2: ..25 322. «Sam 2: 5— MN

165

 

 

 

 

 

 

 

 

"—30

 

 

 

 

 

 

 

 

":50 3 2.8 3 on no» .5 .3
on 2. a e .96
3 on 8» .5 .MN E 8>
E 35. 28 ”Snag E .28 E 02
E 92 E >822 E 8»
E 02 E 8% E 02
$32 uo>oEE_ 758
$3.8 22>» BEEE 98 >2: ““0282 $60.8 0382 2.038 wEBEw 5.: 3:20 swam. .mN
.: 8:25» SEE 335:: 303 2:03 so» 2: Swo— =o> En 80:3 62 E58 0382 8382 .55 8:223 3:2:
no.3 £5 5 22:8: x55 so» on— .30 :0 090 8:3 .oaO 88m .30 :0» BO .aaO no» 26: $0 33295
:2“ .a0
ﬁssazo 20:56
20?ch «8325
825820 «:5QO
20.850 2 m. 8

 

 

 

 

 

166

U

DO

U

m
02925

D D D D
D U D D
D D D D
U U D D
D D D D
D D D D
D D D D
v m N _
35¢

bwﬁraw 3525 .325 Z 02w< buﬁrzm

.=2 8 50.3 m: 2 wow 2:2: 355:: .5 :20 Am
22:30 35:.

:30. :9: 33388:: 22: 3255 2 wow 8:2: 355:: C
.3_::§:_:U 35::

:30— :S: 53388:: 20E 3255 Snow 35:: 355:: Av

.585 o: 53 2: M: mom 3:5 355:: .5 53m A:

$558 5.: 30w : m:m0m but? 052: 3555: .:o :35 2:. Au
L2:355

:3: 5.: 30w 2: m— mUm b03> 3:“:— 3555 .:o 52w 2:. 3
.3ow

328 m— mUm burg 35:: 355:: :o 03:: «353:: 52:55 A:

 

3m Um 3:2: 35.. E: 25:: 2:23:5m

.75: 2:55.55: Mum—:0 2:2 :3 3:23:52 Gamay:

MN 535.5. 3:33: 5 325.5. 3%: 5% 5.3:? 0:. :2 235 .3552: 5 3ng so» :: 555:2 :03 5:3 3:3: 5 356.3» 32%:

5» 552—3 236.: :3 £323 2: o: 32: 23.5 2:323 :3 2m Um 33:3 022: 355:: :53 8:23:22 052 32 :3» _ .3 .0

.3335 3:2: 355:: .:o 5:: 5% o: :33 :53 532058 m 05: o: 2:: 2:03 : .252

mmuhmaabv ”EN—<2 95:37:): 230,—. meDH—HH< .mmméh "N ZOF—bﬂw

167

U

m
00555

.m_u::::m 00:50.:—

O

DOC]

D
v

U

DO

U
M

5:57:

U

DO

U

D
N

005:.

U

DO

U

—
00:u<

55:55

O

ODD

wﬁvcso:
:05: 5.: 50m m: 58:35 E055 0:5: 555::: .:o :55 0:5.
.voow

5:5: 255% hot? 55:: 555:: :0 05:: «5:30: 05:55

 

$058 :05 0::: .:o 0:53 a m: m: m0m 055:: 555:: b5 0m:
.m0_:0_5> 55:: .50— 5:: :05: 50:55 : 05:00:

m : m0m 55:: 555::: :55 o: :05: m: : 5.55: 05505 :55
.m:m0m 55:: 555::

5:: :05: 50:55 05.550 055:: 500: .52 m: 5.55: 5:?
.m: m0m 55:: 555::

5:: :05: 50:55 5025820 55:: 502 >5: 3 5.55: 5:?
.m:m0m 0:58 555:: :55 255 :05 m2: :: £05.:
.:o5:::::0w 50w 5: m:m0m 85:: 555::

05.550 55:: :50:

5:: 355 055% 0: 58:5: 05:: m: m: m0m 55:: 555::
5025520 0:55 500—

5:: 0:55 0555 0: 55:5: 05:: m: m: m0m 55:: 555::
.5:

0: 550 m: m: m0m 55:: 555:: .5 300 5:58: 0:5: :55

:Eamsm 55:: 55: E: 55:: 050.55%

A:

:0

5

A0

AS
a
e.
A:

c

A:

168

ODD

DD

ODD

ODD

D

U

DO

DOC]

DU

DOD

ODD

DU

DUO

[30

U

DC]

DC]

.338 :3 0:5 .3 83>» a m: «Esmmsm 38:. 335:: .3 83

.333; 38:. 83— 35 33: 30:35 : 3.83: «Esmmsm

E

38:. 335:: :85 o. 33: m: : 3.38. 33:33 5:? A3

.3535 38:. 335::

.85 33: 30:35 03.330 38:. 83— >3— : 3.38.. 3:? 9:.

...Ezmmsm 38:. 335::

:85 3.3: 32.35 33.8820 38:. 83— >3— : 3.38.. 3:? Go

.3533 38:. 335:: 325 E35 3:. :5 E 20:5.“— 3:

.:o_8:_:...ow 30w mm: «8.526 38:. 335:: 88

33.630 38:. 802

:9: $35 380% 9 38:3: 20:. : «Esmmzm 38:. 335::
30:38:30 38:. 802

:85 835 388m 8 3863.. 80:. m: «833m 38.: 335::
:3.

0. $3 m_ «Ezmmsm 38:. 335:: 3 33 3:30: 38:. :3...—
.=3 8 30.3 m: «8.525 38:. 335:: .3 :80

63.830 38:.

88. :85 33368:: 03:. 3235 «8:35 38.: 335::

8:39:50 38:.
802 :85 58388:: 03:. 3235 2:335 38:. 335::
.:8:3 9 $3 08 «5:35 388 335:: .3 33m

w:==:. .8 3w m: 88:35 burg 38:. 335:: .3 :8..w 2:.

3

A»

Q

A3

A>

A:
c

169

 

a. 8.585

 

S 2......
3 $52

 

 

 

 

 

 

 

W

 

 

3.039 a. .550
a. .858
.. 3.039 a: 228 E 8:85
a: 33w .8 23¢
A8 8335 E .82. 35:3...
E 2...... E 888
:8 .82. 35:35 5 282. 3:25
a. 088m 5 28...._oz
E 90...... :25... E .98... .83
a. 23.9.2 3.8 32:. ﬂ
3. 3...... :83
E .25 >5. «33:30:.
.230 mum—Elm— .” 7.20 so» m.3.__3.. 2.. 8:0 3
38.32.. .5. a. a» 8: E n. so... .52. so. ..... ow 8: 5 ﬂ
205...... 2.. 8.. 8» ..... 22.3 .35 8. oz 3 8.» 52.; so... .35 8. oz 3 .3
H2335:

0:. amm— MNE3=E£ 333—3.: 05 :08 33 so» 03: .N—O

 

75...... 2.. :55: ..8._.......
w.:>3=o.: 0:. 30.8 :80: 33 3% 03: .30

 

 

 

 

 

...23335 38:. .8 3: so» .3283.“ .3 09$ 3:. 30:8 .3.» :3: :82 o. 8:: 2:03 : >32

A<H¢DDEZV mmN—A—Hmm—h HAND—SHED EC ZO—HmOA—d. G72 HGGHQBOZv— "m ZO—PUmm

170

.30 bu: :2. «on a .20 can 3.0 .30 3 an 052: :o 355:8 com—ans 38.3.: 53.nov, an... 2: E a: ﬂ

 

 

 

 

 

 

 

 

 

manage.
3.083 E 850 .25
5 28:3 .o£o=m ES. 8365.". g 85
Q .28.: 9.. 8o... 83:25 a. coamsczmm .33 «.58.: oz.» :8 viz
E 6%.. 5558. 9.. so... 388.5 E as: 9555 on. .< 8. 229: no:
3.8 mmhza MN 8. 9553 98mm 3 829: c9620
E... :52... ﬂ 3. 22....
7...... «5.7.... ﬂ
$253.88 2.28358 so» SEEN: 05 Dana .30»
:0» SEE»: 0:. 2.3.8 :9» 2.. 20:3 :20 2.. 5:3 £83823 wEEE: 0:. wESQ £20 SEES": .20

 

.33 £5 c.599— ...ou_=t£ 9:. .3 «Eu: 0:. 3:825.— 25 33. 38.5.: .5 MN

@538 9:38» am..— oﬁ E 33:. co bag“ so» 2:. Eoﬂztom 3:3 .3 .0

171

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6 88b 85
3 825,.— 8 e viz
>
A 52.3 E .25
A Await *5
N 29: ﬁg
3 ﬂags =65
9 co
_ a mmth ﬂ 8 cu 2. E ﬂ .2 - £0 8 on 8:: ﬂ
$52. so
203:5 305 33% as: 33 $3.0 § 02 3 mi
7.5:... 2.. 93E. ﬂ
~35: :o Eoﬂztﬂ
w£>>o=£ 05 3:93 35 so» 963 .mﬁO
v
m
N
_
”AEE .28
3 $5
5 E2
5 x .82 S 222. 8:
E x .82 E 2252 5620
E 65 .2 § 2222
.23.? 53: ﬂ .38 maze. MN .352
.ﬁEa 525 Seaman. 5:0 Egon .310 8:

 

 

 

735 2.. 5 5:: ﬂ 23:20.“ 3382 as 383 so» Mo 52803 as g 3:3 .2 .0

172

 

 

 

 

 

E 255. 28

 

 

 

 

 

 

 

E m; ”5683:9550
3:2 3% Easiﬁ
$33 QB: :20 3 a» vi E § 02
.23 323.5 0.8 >05 .2 cote—60E $382 no» 0332 so» ARM.“ «BE—En:
BEE 5m Sﬁztom Hows 2:03 Ev .80. mo ab Ev 80:3 2038 wEBEw «as «620 you 5.0.8 3332 £50
33 £5 E 2055 x55 :9» cm .150 “a; 6:0 88m 550 0208.. :9» 29 £50 35 so» 032 .50 BEES”—
...a .20

 

 

was» 805 5 BE: :0 uni—Eu.“ warm—nan mom so» En >55 .930

"7:583. w£>$=£ 2: in 6.3 out“... .3 uni—3.5.. wﬁbnnn 3.53.». 38.3.“ 3.3» 2: ..85 322. 35% 2: 5 ﬂu

 

5 A23: 85

 

8 $8.59. .5 e E2

 

 

 

 

we we 8 mo

 

5

co 3

 

 

mo

 

 

no

co

 

 

 

€ 8> 2: 02
was?» 806 mats—v BEE :o 5E=t8 bag so» 20 £30

 

co ..uﬂztom wﬁbanm

.oNMaE

 

83% 30> 35 §=§ 32526

 

 

$338 so “35:0.“ wﬁbaam voqmoum .8 95.8% so» 35m mag 38% >58 30$ 37.: .39» 338 Loan—«E9. «a: 83. 35.3.. a: MN

.23 2.. a 3::

MN 63:2: co ..oNEtom 32805 warn—ans gm :0» Eu 5:? A. .. a3 .0

173

D D D D U do; BEE 8m voow m_ we: .8:th Aoo

w v m N _
93935 3uu<

bunobm 09.935 1.5.52 3.3.. bun—Pam

 

c oh. .5 «2:33:98.— uacaa ﬁaoﬁsaym

 

D D D D D 2288 was 2:: go 833 a m_ 3.52.;2 Bﬂztom Ac:
.mcoEvcoo

D D D D D 5583 8 26 225 no.5 mo 8:83 but ﬂ VEZ had—Eon— ASE
D D D D 0 £8 2: .8 voow go: a VEZ Bﬁztom E
D D D D D .ozmaaxo .4 E2 55:3; 3
D D D D D .538 2 $8 m_ Vaz 3N=E£ 3
D D D D D .388 :o vEZ bang 2:05 88 ma: 5 825$ E
D D D D D .22» as... 3:205 E2 .o~_=c£ :2
D D D O D dob BEE ..om uoow ﬂ v52 .oﬁztom Aww
w v n N —

0293.5 09.?

bwucbm 92»:me 3.5.52 oouw< bunchm Q .5 v52 33%

.78: Sara—Emma MUHEU .2an can Suva—83m
gum—y: ﬂ 336$. 3.5%ch 8 bwzchm 8&3 so» .052—3 2: :8 333 {Swag ho 3&3 so.» : .EoEBSm some .33 ~33»: 5 «3&qu
69%» so» 850:3 88:2: 28 £388 0E 8 com: 885 .85 28 v5 Z EMS—End .8155 Sons mucoEBSm 2:8 32 EB. .2 .0

dob SEE 89m so 33 98 vEZ Eoﬁ—Eom 33805 9.31330 x35 so» ow 8:3 Sons :ozmmbéoo a 26: 8 BE 2:03 _ .30 Z

AdammDU—mzv mmN—A—szh A<U=ZH=U 339—. man—PEPE! .me2~—<h "v ZOE—9mm

174

 

.336

D D D :o 38: :5 vEZ 83:33.: 32:55 @3153 :8 2:: of. G
.323:
D D D :o 32: :5 VEZ £8235: 32523 wﬁbnqa 5m 3052: 2:. Ac
D O D .338 :0 bag: 8 SEE»: 33:23 no cab 2B. 3
D D D .808 33:: 3.3895 no 83:33:36 :5 mowﬁﬁgum 2C. 3
oEaau—STREM u_a3w_uo_3::v— «Ewen—5.320—
ucz «2.32:5 53> a
72.

8338.5: 2: 3923 .5: 3:06:38: 3 039m 2: 6:: 8E3 2: am”: MN .838 085 :53 03832305 38 so» 0: 3on

3:3 9 2: =3 23 353.30 :03: 885 .323 93 v52 Boﬁztﬂ 33823 :5 motor? 33:: 333:: :o 389 2:8 :32 EB _ 302 .35

30:08 :5 2:: mo 883 a m_ no.5 :03:th As:
.=o.u. 05 :8 voow Ho: .9: 3.5 hoﬁzton— 9
632098 ﬂ «2: :3:th 3m

:33: 8 >30 .0: 32: .53th"— A:

.33.: :o «8: bag“ 2:05 33 £5 E ﬂogm 30

000000
[300000

[300000
UDDDDD
000000

.203 33:. 88203 32: :oﬁztom a:

175

33% £5 van—:88: 23% 30:8 >381: voﬁiEoU
33% 5E E3. 03% :95 :02:on $33 83 _ 23:» 30:8 .0383: 080w
32:3 :0 30: 8:53 883::

3:56:38:

 

3 32:283.: cue: P.ZOG 3:9? :2: :3 302—0— .” mEmEc “88:8 283 30% E: 8:38 :o 333 :ms 2.: m: 355 ”mud

 

 

“9:8va :oﬁO

 

:m_> COECDHXW

 

2:23:

 

w:_:3::.

 

 

 

 

@8353? on: :0 88:8 2: 2:53: :o» :8 832m .83 .0 $538: :0» E: 83633 53:25 :0 090 3:3 .3.N~ .O

 

TBS 2: 5 H1::— .m?m hi MN D :2 D 99> «£808 N: :3. of :8 383233 83:35 338: so» 20 .NN .O

 

 

$23383
2:3 322:8 :38 83 2:3: 2:: a 3:3 "30> .5 MN .8: .0

0 oz 0 8> 22.2083 .2083: .8 :BEoE n so» o:< AN .0
A: :35 208v n 29 3:38.: A3

I 9:8: 2.2: :33. 3:

“go: 3 Es: 5:22: 3

.0300: | ““2283; m3: 8 o>= 2:8: >82: >62 .3 .0

8:8 :5 383: SPA mo 88%: 338% 893 :32 8 BE 2:95 H 538::

mU—ln—EUOEHG Q72 ZO—HAQAM—Ormzu dém—Zm—U "m ZO—Humw

176

0:003:00. :88080 .0:=:_=2%0 .3 80803.88: 0:: 0: :0_:=:_:::00 23% 0 0008 02:: 30> 08:: :50» :0: :0» 0309—.

 

 

:08 ::_3 0:0:m 0: 00:— 01.03 :0» :0:: .00:0 2:: 8 50300:: 0208 80:0 808800 :0 380:: 80 0>0: :0» 00 .36

0:080:

202

 

3:505:00: 0: 8038—8020 000: EEOG 023:0 :05 :0 200:0. MN H:00000 ﬁnd

.380 A: :0 200 :0 :25 0 :00» 0:: :85 6%: 0 : :0 8080: : :00 — .800.» :0w0 :00» :0: 0802: so» :00 .35
u . .: : : . .: .MN

38:00 38:03:: 00:2:808 0%230 00:2:800
3:838: 000:0:08 0%230 080m III

300% 5:03: 00:2:808 30:8 :wE 00:2:800 I

300% ::.:_03: 0:0 000% :::0>8 503:0: 820 88 30:8 :wE 080m III
300% :88 00:29:08 30:8 b00308 00:2:800 II

300% :::0>8 0:0 000% :5: 503:0: 820 808 30:8 b00308 080m l
300% :::0>8 00:2:808 = 000% 30:8 E080: 00:2:800 II

300% :80>8 0:0 000% 8%.: 503:0: 820 808 = 000% 30:8 .0081: 080m

177

Appendix E: Exploratmy and Conﬁrmatmy Factor Analysis

 

178

Exploratory and conﬁrmatory factor analysis
The decision whether items should be combined into a scale was made on the
basis of results of exploratory factor analysis, with maximum likelihood procedure and
oblique rotation. Items were combined if they had high loadings on the same factor and
suggested a fairly similar concept. To judge the appropriateness of the number of factors,
we combined the interpretation of substantive meaning of the factor analytic model with
three distinct statistical criteria:
- The Chi-square test, at a = .05, which tests if the number of factors in the solution
ﬁts perfectly in the population.
- Test based on eingenvalues. For the solution to ﬁt perfectly in the population, the
number of factors should be equal to the eigenvalues that are greater than 1.
- Residual correlations: For the solution to ﬁt perfectly in the population, residual
correlations with absolute values greater than .05 should be less than 20 percent.
C onﬁrmatory factor analysis
A conﬁrmatory factor analysis (CFA) was performed to conﬁrm the factor model
obtained with exploratory factory analysis (Borjesson et al. , 2003; Grimbeek and Nisbet,
2006). The following ﬁt indexes (Kline, 2005) were used:

- The Chi-square statistics (CMIN). This tests the null hypothesis (Ho) that the
model ﬁts perfectly in the population. A small value of CMIN allows us not
to reject the null hypothesis.

- The root mean square error of approximation (RMSEA). This tests the null
hypothesis (Ho) that the model has a close approximation ﬁt in the population.

A rule of thumb is that RMSEA s .05 indicates a close approximate ﬁt, values

179

between .05 and .08 suggest a reasonable error of approximation, and RMSEA
2.10 suggests poor ﬁt.

The comparative ﬁt index (CF I). This assesses the relative improvement in ﬁt
of the model compared with a baseline model. A rule of thumb is that values

below .90 indicate poor ﬁt, and values above .95 indicate a good ﬁt (Kline,

2005)

180

F_igure 2: CFA model for SC513.

  
 

Attitude toward
marketability of improved
maize variety SC513

.73

Attitudes toward production
and consumption traits of

improved maize variety

SC513

     
   

 

   
   
    
  
   

.58

 

Grain from SC513 is
easy to sell.

 

 

.35

 

Fresh maize SC513 is easy to
sell.

 

 

 

.25

 

Cultivation of SC513 is a waste
of time and money.

 

 

Two-factor model for SC513 (standardized estimates)

 

Seed from SC513 has good
germination.

 

 

 

 

 

Grain from SC 513 is good for
milling.

 

 

.16

 

 

When rainfall is low, Chimanhica has
better production than SC 513.

<—- udroug

 

 

 

181

 

 

P_‘_igure 3: CFA model for NPK.

 

One-factor solution for NPK (standardized estimates)

.72
NPK is good for maize.

 

 

 

 

 

 

   

 

 

 

 

 

.85 .49
NPK increases yield.
. .19
Attitudes toward NPK ,
‘; NPK is a waste of time and
.16 mom... a-
03

 

NPK is harmful for the soil. .

 

 

 

 

" 'o" -"I-l'l

 

 

 

[figure 3: CFA model for Urea.

 

One - factor solution for urea (standardized estimates)

.92
Urea is good for maize.

 

 

 

 

 

 

 

 

 

 

 

.96 .77

Urea increases yield.
.88
. .15
Attitudes toward urea .38
_; Urea is a waste of time and

.29 money. h-

09

 

Urea is harmful for the soil. h.

 

 

 

 

 

 

182

Appendix F: Collinearity Diagnostics and Logistic Regression

 

183

FOVé

”Fm 0w Uﬂucma L0>m 6.05.5) “aways—800 .m

mFmom 05:3 835m

 

184

 

 

3F. 8o. 8? Ba. F8. new. new «a? 5:092...
mmFF 5m. «8. Ea- m2. 80. 2a.- 8.683920 mewzw
83 8F. 80. 83- 9a.- wmo. m3- 2me 3:8 83?.
«8. F N Fm. one. 83- SF: 08. ET “9.55 F.
5 F .N 8v. 3 F. 5.. F- t F.- m8. 8a.- 2:892 F.
mmFF 8w. F Fe. 23 8F. m8. omm. 88.205. 55.5
> Em 30> c 2w
9.3 80. at. F Fm. 5 5. Bo. moo. 28th “Ems r6...
mmFF m8. F Fe. 8m. 95. m8. So. 8:83.... Ease
FvFF o3. m8. m8. F- 08.- N8. «8.- 28:88. Fo mm<
mm
ommF 3K. o8. 83 own. So. am. comma mnﬂcwwhwhwm
F8. 30m 2:. «8. 98569
“.5 8:9.an gm F Sam 55 .3 m .822
8%:me 3:35:00 956580 9565000
352%:me 8523:923
!
3535000

.mFmUm but? 388 BREE: mo gouge?“ :o commmocwoc ocmmwn: F8 @835 813:3 ”F 2%;

 

moumocuﬂw 3503on

 

mums. :0 x12 now: .96 uo_nm_.m> Fcoucoaoo .m

 

 

 

oooF oFo. moo. NSF- ooF.- ooo. ooo.- ooumodwwwmm
Fom. F ooo. ooo. mono ooo. ooo. mom. 8:952; 3&8
BF. F woo. Foo. omF. ooo. ooo. ooo. 8:55.85 5&3
93 3o. ooo. on: ma. Foo. mom. 88.35 5&8
o FNF 8o. ooo. NE. m3. ooo. ovo. 8:838 Egon.
. . . . . . . >=Em¢ 50> 5 2m
Ro F Eo omo Foo moo So moo 3855 2.2. so...
ovFF FE. ooo. Roy ooo.- woo. ooo. 28832 oo oo<
FoFF Foo. Foo. Food ooF. ooo. «NF. 68 553
o Fo. ooo.- 2 F. ooo.- 98388 F
".5 8:228 .o_m 2 Sam BE .on m .282
83:3 €85.60 $56580 386580

352335 umﬁEmucﬁmc:

3:33:30

.VEZ mo corneas so ~868ch otmmwﬁ How @235 353:; ”m 2an

185

 

 

0.20.: :0 005 000: :0>m ”0_nmtm> 2.005000 .0

amt-d U.— “30¢

 

 

 

oNNF oFo. Foo. So. F- 8.. ooo. ooo.- 883:? 8:80
83 o2. ooo. R2 3:. NB. Fom. 80:93 5:50
EFF ooo. ooo. 05.- ooo.- moo. «8.- 09:8. 5:50
R3 8:. ooo. oooo ooo. Foo. oFo. 88.382 5:50
SS o5. oR. mom- oFo.- Foo. Noo- awhmmw “$35..”
ova So. 8:. own. so. ooo. moo. 8:838 5:50
FoF.F Foo. :8. Non.- o3.- moo. 6? E8882. .0 8<
ooFF ooo. ooo. 85. NR. ooo. mom. :2 52:0
000 ooo. oNF. ooo. 0:80:80 F
-.=> 8:228 F .28 BE .90 m .88.
8:250 3:82.00 280580 2:20:88
0020:0035. 00N_0._m0cmch3
«E20580

.008 m0 0000000 :0 0082?: 000%2 00.“ 08800 833.03 ”m 053.

186

Meagres of ﬁt in logistic

Proportion of reduction error using the model
R2 Logistic = [(-2LLo)-(-2LLM)1 / <-2LLo)

SC513
R2 Logistic = [(-2LLo)-(-2LLM)] /('2LL0)
R2 Logistic = (224.28 — 137.26) / (224.28) = 0.388

NPK
R2 Log-sti- = [(-2LLo)-(-2LLM)] / (-2LLo>
R2 Logistic = (264.94 - 200.60) / (264.94 ) = 0.243

Urea
R2 Log-sti- = [(-2LL->-(-2LLM>1 / (-2LLo>
R2 Logistic = (300.35 - 218.77)/ (300.35) = 0.272

Computations of (some) predicted probaglities

1: = odds / (1 +odds)

The logistic equation (model) for improved maize variety SC513
Log (rt/l - 1t) = a + BIAEZ + BzAGE + B3FAM + B4EDU + BsKNOW + BNNEIG +
BMMARK + BéATTmark + B7ATTchar + BgADP*ATTmark

Log (rt/l - 1t) = -.704 + 2.51AEZ -.020AGE + .005FAM + .745EDU + 1.99KNOW -
.745NEIG -1.15MARK -.497ATTmark -.8l2ATTchar + .815ADP*ATTmark

Predicted probability of adoption of SC513 for someone in highlands having some
education and being knowledgeable of advantages and disadvantages of improved maize:
Log (ﬂ/l - 1t) = -.704 + 2.51 -.020 + .005 + .745 + 1.99 -.745-(1.15*2) -.497 -.812 +

.815= 0.987

Log (7t/1 - 1t) = 0.987

Odds of adopting = eA 0.987: 2.68

Probability of adopting = 2.68/ (1 +2.68) = 0.728

187

" “7'!

 

Predicted probability of adoption of SC513 for someone in lowlands having some
education and being knowledgeable of advantages and disadvantages of improved maize:
1t = odds / (1 +odds)

Log (rt/1 - 1:) = -.704 + O -.020 + .005 + .745 + 1.99 -.745-(1.15*2) -.497 -.812 + 0 = -
2.338

Odds of adopting = e" -2.338 =.0965
Probability of adopting =.0965 / (1+.0965) = 0.088

Predicted probability of adoption of SC513 for someone in highlands having some
education and being knowledgeable of advantages and disadvantages of improved maize,
who learned about improved maize from neighbors:

Log (712/1 - 7t) = -.704 + 2.51 -.020 + .005 + .745 + 1.99 -.745-(1.15*0) -.497 -.812 + .815
=3.287

Odds of adopting = e" 3.287 = 26.76

Probability of adopting =26.76 / (1+26.76) = 0.964

Predicted probability of adoption of SC513 for someone in highlands, having some
education and being knowledgeable of advantages and disadvantages of improved maize,
who learned about improved maize from market.

Log (rt/1 - 1t) = -.704 + 2.51 -.020 + .005 + .745 + 1.99 -0-(1.15*2) -.497 -.812 +

.815=] .732

Odds of adopting = e" 1.732 = 5.65

Probability of adopting = 5.65 / (1+ 5.65) = 0.850

Predicted probability of adoption of SC513 for someone in highlands having some
education and being knowledgeable of advantages and disadvantages of improved maize,
who learned about improved maize from extension:

Log (ﬂ/l - 1t) = -.704 + 2.51 -.020 + .005 + .745 + 1.99 -0-0 -.497 -.812 + .815=4.032
Odds of adopting = e" 4.032 = 56.37

Probability of adopting = 56.37/ (1+ 56.37) = 0.983

Predicted probability of adoption of SC513 for someone in highlands having some
education and being knowledgeable of advantages and disadvantages of improved maize
and attitude score = 1:

Log (“It/l - H) = -.704 + 2.51 -.020 + .005 + .745 + 1.99 —.745-(1.15*2) -.497 -.812 +
.815= 0.987

Log (1t/l - 1t) = 0.987

Odds of adopting = e" 0.987= 2.68

Probability of adopting = 2.68/ (1 +2.68) = 0.728

Predicted probability of adoption of SC513 for someone in highlands having some
education and being knowledgeable of advantages and disadvantages of improved maize
and attitude score = 2:

Log (ﬂ/l - 7t) = -.704 + 2.51 -.020 + .005 + .745 + 1.99 -.745-(1.15*2) - (.497*2)-
(.812*2) + .815= -0.322

188

 

Odds of adopting = e" -0.322 = .725
Probability of adopting =.725 /(l +.725) = 0.420

The logistic equation (model) for N_P_K_

Log (rt/l - n) = a + [31ADP + BzAGE + B3FAM + B4EDU + BSKNOW + BMMARK +
BEEXT + BéATTnpk

Log (rt/1 - n) = -4.32 + .794ADP -.001AGE + .041FAM + .246EDU + 2.61KNOW +
.154MARK + 1.57EXT -.322ATTnpk

Knowledgeable

Log (1t/l - 1t) = -4.32 + .794 -.001 + .041 + .246 + 2.61 + .154 + (1.57*2) -.322 =2.342
Odds of adopting = e" 2.342 =10.4

Probability of adopting = 10.4/ (1 + 10.4) = 0.912 I:

Not knowledgeable

Log (rt/l — 1t) = -4.32 + .794 --001 + .041 + .246 + 0 + .154 + (1.57*2) -.322 = -0.268
Odds of adopting = e" -0.268 = .7649

Probability of adopting = .7649 / (1 +.7649) = 0.433

 

The logistic equation (model) for ure_a F
Log (rt/l - n) = a + BIAEZ + (3on13 + BgFAM + B4EDU + (35KNow + BMMARK + -'
BgEXT + BéATTurea

Log (rt/1 - 1:) = -2.86 + 1.43AEZ - .013AGE - .018FAM + .256EDU + 2.05KNOW -

.211MARK + 1.23EXT -.251 ATTurea

Knowledgeable

Log (rt/1 - it) = -2.86 +1.43-.013 - .018 + .256 + 2.05 - .211 + (1.23*2) -.251= 2.843

Odds of adopting = e" 2.843 = 17.17

Probability of adopting = 17.17/ (1 +17.17) = 0.945

Not knowledgeable

Log (rt/1 - 1t) = -2.86 +1.43-.013 - .018 + .256 + O -.211+(1.23*2)-.251=0.793

Odds of adopting = e" 0.793 = 2.21

Probability of adopting = 2.21/ (1 + 2.21) = 0.688

The logistic equation (model) for MK;

Log (rt/1 - n) = or + BlADP + BZAGE + [33FAM + B4EDU + BsKNOW + BMMARK +
BEEXT + [36ATTnpk

Log (n/l - 1:) = -4.32 + .794ADP -.001AGE + .041FAM + .246EDU + 2.61KNOW +
.154MARK + 1.57EXT -.322ATTnpk

Machipanda
Log (rt/1 - ‘11) = -4.32 + .794 -.001 + .041 + .246 + 2.61 + .154 + (1.57*2) -.322 =2.342

Odds of adopting = e" 2.342 =10.4
Probability of adopting = 10.4/ (1 + 10.4) = 0.912

Vanduzi
Log (1t/l — 1t) = -4.32 + 0 -.001 + .041 + .246 + 2.61 + .154 + (1.57*2) -.322 =1.548

189

Odds of adopting = e" 1.548 = 4.7
Probability of adopting = 4.7/(l +4.7) = 0.823

The logistic equation (model) for m

Log (713/1 - 712) = (I ‘1' BlAEZ + BzAGE + B3FAM + B4EDU + BsKNOW + BmMARK +
BEEXT + BgA'ITurea

Log (rt/1 - 1t) = -2.86 + 1.43AEZ - .013AGE - .018FAM + .256EDU + 2.05KNOW -
.211MARK + 1.23EXT -.251 ATTurea

Machipanda
Log (rt/l - 1!) = -2.86 +1.43-.013 - .018 + .256 + 2.05 -.211+(1.23*2)-.251= 2.843
Odds of adopting = e" 2.843 = 17.17 3

Probability of adopting = 17.17/ (1 +17.l7) = 0.945

Vanduzi

Log (1t/1 - 1t) = -2.86 + 0 - .013 - .018 + .256 + 2.05 -.211+(1.23*2)-.251=1.413
Odds of adopting = e" 1.413 = 4.11

Probability of adopting = 4.11/ (1 + 4.11) = 0.804

_ll‘l\ '1'“.-

 

The logistic equation (model) for NPK
Log (TC/1 - it) = -4.32 + .794ADP -.001AGE + .041FAM + .246EDU + 2.61KNOW +
.154MARK + 1.57EXT -.322ATTnpk

Extension

Log (rt/1 - 7t) = -4.32 + .794 -.001 + .041 + .246 + 2.61 + O + (1.57*2) -.322 =2.188
Odds of adopting = e"2.188= 8.92

Probability of adopting = 8.92/ (1 + 8.92) = 0.899

Neighbors

Log (rt/l - n) = -4.32 + .794 -.001 + .041 + .246 + 2.61 + O + 0 -.322 = -0.952
Odds of adopting = e" -0.952= .386

Probability of adopting =.386 / (1 +3 86) = 0.278

The logistic equation (model) for m
Log (n/l - n) = -2.86 + 1.43AEZ - .013AGE - .018FAM + .256EDU + 2.05KNOW -
.211MARK + 1.23EXT -.251 ATTurea

Extension

Log (“It/1 - 1!) = -2.86 +1.43 - .013 - .018 + .256 + 2.05 - 0+ (1.23*2) -.251 =3.054
Odds of adopting = e" 3.054= 21.19

Probability of adopting = 21 .19/ (1 + 21 .19) = 0.955

Neighbors

Log (rt/l - n) = -2.86 + 1.43 - .013 - .018 + .256 + 2.05 - 0+ 0 -.251 =0.594
Odds of adopting = e" 0.594 = 1.811

Probability of adopting = 1.811/(1 +1.811) =0-644

190

REFERENCES

Abebaw, D., & Belay, K. (2001). Factors inﬂuencing adoption of high yielding maize
varieties in Southwestern Ethiopia: An application of logit. Quarterly Journal of
International Agriculture, 40, (2), 149-167.

Adesina, A. A., & Zinnah, M. M. (1993). Technology characteristics, farmers’
perceptions and adoption decisions: A Tobit model application in Sierra Leone.
Journal of Agricultural Economics, 9, 297-311.

Aiken, L. R. (2002). Attitudes and related psychosocial constructs. Theories, F
assessment, and research. Thousand Oaks: Sage Publications. .-

Alreck, P. L., & Settle, R. B. (2004). The Survey Research Handbook (Third Edition).
New York: McGraw-Hill Irwin.

Anderson, R. J ., & Feder, G. (2004). Agricultural extension: Good intentions and hard .‘i
realities. The World Bank Research Observer, 9(1), 40-60. ;

 

Babbie, E. (1998). The practice of social research. Belmont, CA: Wadsworth
Publishing Company.

Beets, C. W. (1990). Raising and sustaining productivity of smallholder farming
systems in the tropics. A handbook of sustainable agricultural development.
Alkmaar, Holland: AgBé Publishing.

Belay, A., Claasens, A. S., & Wehner, F. C. (2002). Effect of direct nitrogen and
potassium and residual phosphorus fertilizers on soil chemical properties, microbial
components and maize yield under long term crop rotation. Biol F ertil Soils, 35,

240-427.

Bias, C., & Donovan, C. (2003). Gaps and opportunities for agricultural sector
development in Mozambique. Research Paper 54E. East Lansing, Michigan:
Michigan State University, Department of Agricultural Economics.

Borjesson, W. I., Aarons, G. A., & Dunn, M. E. (2003). Development and 'conﬁrmatory
factor analysis of the abuse within intimate relationships scale. Journal of
Interpersonal Violence, 18, (3), 295-309.

Byerlee, D., & de Polanco, H. E. (1986). Farmers’ stepwise adoption of technological
packages: Evidence from the Mexican Altiplano. American Journal of Agricultural
Economics, 68, 519-527.

191

Byerlee, D. & Heisey, P. W. (1997). Evolution of the African Maize Economy. In
Byerlee, D., & C. K. Eicher (Eds), Africa ’s emerging maize revolution (pp. 9-22).
Colorado: Lynne Rienner Publishers.

Byerlee, D. (1994). Maize research in Sub-Saharan Africa: an overview of past impacts
and ﬁtture prospects. Economics, Working Paper 94-03. Mexico, D. F.:
CIMMYT.

CIMMYT (International Maize and Wheat Improvement Center). (1993). The adoption
of agricultural technology. A guide for survey design. Mexico, D. F.: CIMMYT.

Chatterjee, S., Hadi, A. S., & Price, B. (2000). Regression Analysis by Example (Third
Edition). New York: Wiley.

Chilonda, P., & Van Huylenbroeck, G. (2001). Attitude towards and uptake of
veterinary services by small-scale cattle farmers in Eastern province Zambia.
Outlook on Agriculture, 30, (3), 231-218.

Denic, M. (2005). Biotechnology, breeding and system for African crops. New OPV
varieties released by maize Team of INIA. Retrieved May 20, 2005 from
http://www.africancrops.net/Research/fato/outputs.htm

Diallo, I. (1983). An assessment of farmers’ attitudes and experience under no—till corn
production system in Baltimore County, Maryland. Unpublished doctoral
dissertation. University of Maryland.

DNER (National Directorate for Rural Extension). (2002). Internal Review
DNER/SG2000 Mozambique. Unpublished mission report. Maputo, Republic of
Mozambique: Ministry of Agriculture and Rural Development.

Eagly, A., & Chaiken, S. (1993). The psychology of attitudes. Belmont, CA:
Wadsworth Publishing Company.

Easterly, W. (2006). The White Man’s Burden. Why the West 's Eﬂorts To Aid The Rest
Have Done So Much Ill and So Little Good. New York: The Penguin Press.

Eicher, C. K, & Kupfuma, J. D. T. (1997). Zimbabwe’s Emerging Maize Revolution. In
D. Byerlee, & Eicher, C. K. (Eds), Aﬁica ’s emerging maize revolution (pp. 25-43).
Colorado: Lynne Rienner Publishers.

Enyong, L. A., Debrah, S. K., & Bationo, A. (1999). Farmers’ perceptions and attitudes

toward introduced soil-fertility enhancing technologies in Western Aﬁica. Nutrient
Cycling in Agroecosystems, 53, 177-187.

192

 

 

 

Feder, G., Just, R. E., & Zilberman, D. . (1985). Adoption of agricultural innovation in
developing countries: A Survey. Economic Development and Cultural Change, 33,
(2), 225-298.

Fumo, E. (1995). (Analise dos factors que inﬂuenciam a preferéncia das variedades de
milho e o seu sistema de cultivo no sector familiar do Distrito de Manica).
[Analysis of factors inﬂuencing maize variety preference in the Manica district].
Unpublished dissertation thesis. Eduardo Mondlane University. Maputo.

Ganpat, W. G., & Bholasingh, D. (1999). Attitudes of farmers toward farming in
Trinidad. Journal of International Agricultural and Extension Education, 6, (3),

33-38. a

Gemo, H., Eicher, K.C., & Teclemariam, S. (2005). Mozambique ’s experience in
building a national extension system. East Lansing, Michigan: Michigan State
University.

Grimbeek, P., & Nisbet, S. (2006). Surveying Primary Teachers about Compulsory
Numeracy Testing: Combining Factor Analysis with Rasch Analysis. Mathematics is
Education Research Journal, 18, (2), 27-39. "

 

Hair, J. F., Black, W. C., Babin, B. J ., Anderson, R. E., & Tatham, R. L. (2005).
Multivariate Data Analysis (Sixth Edition). Upper Saddle River, New Jersey:
Pearson Prentice Hall.

Hemmerlgarn, A. L., James, L. R., Mitchell, T. R., & Ladd, R. T. (1995). Testing for
cross-situational-consistency: A conﬁrmatory factor analytic approach. Journal of
Management, 21, (1), 121-139.

Hintze, L.H. (2002). Variety characteristics, transactions costs and
maize adoption in Honduras. Ph.D. Dissertation. North Carolina State University.
Raleigh, NC.

Hintze, L. H., Renkow, M., & Sain, G. (2003). Variety characteristics and maize
adoption in Honduras. Journal of Agricultural Economics, 29, 307-317.

Hoeft, R. G-, Nafziger, E. D., Johnson, R. R., & Aldrich, S. R. (2000). Modern Corn
and Soybean Production (Sixth Edition). Champaign, ILL: MCSP Publications.

Hosmer, D., & Lemeshow, S. (2000). Applied Logistic Regression (Third Edition).
New York: Wiley.

Huffman, W. E. (1974). Decision making: The role of education. American Journal of
Agricultural Economics, 56, (1), 85-97.

193

ICRISAT (International Crops Research Institute for the Semi-Arid Tropics). (2004).
Perﬁl de Agricultura Local do Distrito de Manica. Maputo, Mozambique:
ICRISAT.

INE (Instituto Nacional de Estatistica) [National Institute of Statistics]. (2001). Censo
Agropecuario. [Agricultural and Livestock Census]. Maputo, Mozambique: INE.

INE (Instituto Nacional de Estatistica) [National Institute of Statistics]. (1997). Census
97. Maputo, Mozambique: IN E.

INIA (Instituto Nacional de Investigacao Agraria) [National Institute for Agricultural
Research]. (1997). Recomendacoes de adubacao azotada e fosforica para culturas
anuais alimentares e algodao em Mocambique. [Fertilizer Recommendations for 8:
Cotton and Food Crops in Mozambique]. Report No. 88. Maputo, Mozambique:
INIA.

Jeje, J. J ., Machungo, C., Howard, J ., Strasberg, P., Tschirley, D., Crawford, E., &
Weber, M. (1998). What Makes Agricultural lntensiﬁcation Proﬁtable for
Mozambican Smallholders? An Appraisal of the Inputs Subsector and the 1996/97 .
DNER/SG2000 Program. Volume II: Main Report. East Lansing, Michigan: :
Michigan State University, Department of Agricultural Economics.

 

Kaliba, R. M. A., Verkuijl, H., & Mwangi, W. (2000). Factors affecting adoption of
improved maize seeds and use of inorganic fertilizer for maize production in the
intermediate and lowland zones of Tanzania. Journal of A gricultural and Applied
Economics, 32, (1), 35-47.

Kendal, M. G., & Buckland, W. R. (1971). A Dictionary of Statistical Terms (Third
Edition). Edinburgh: Oliver & Boyd.

Kerlinger, F. N. and Howard, L. B. (2000). Foundations of Behavioral Research.
(Fourth Edition). New York: Wadsworth.

Kim, M. S., & Hunter, J. E. (1993). Attitude-behavior relations: A meta-analysis of
attitudinal relevance and topic. Journal of Communication, 43, (1), 101-142.

Kline, R. B. (2005). Principles and Practice of Structural Equation Modeling (Second
Edition). New York: The Guilford Press.

Knoke, D., Bohmstedt, G. W., & Mee. (2002). Statistics for Social Data Analysis
(Fourth edition). Itasca, 111.: F. E. Peacock Publishers.

194

Kumwenda, J. D. T., Waddington, S. R., Snapp, S. S., Jones, R. B, & Blackie, M. J.
(1997). Soil fertility management in Southern Africa. In D. Byerlee, & C. K.
Eicher (Eds), Africa ’3 emerging maize revolution (pp. 157-172). Colorado:
Lynne Rienner Publishers.

Lawal, O. B., Saka, O. J ., Oyegbami, A., & Akintayo, O. I. (2004). Adoption and
performance assessment of improved maize varieties among smallholder farmers in
Southwest Nigeria. Journal of Agricultural & Food Information, 6, (1), 35-47.

Litchtenberg, E., & Zimmerman, R. (1999). Information and farmers’ attitudes about
pesticides, water quality, and related environmental effects. Agriculture,
Ecosystems and Environment, 73, 227-236.

Litwin, M. S. (2003). How to Assess and Interpret Survey Psychometrics (Second
Edition). The Survey Kit 2. Sage Publications.

Massingue, J ., Rafael, G., Haji, U., Low, J ., & Boughton, D. (2004). (Sector Comercial
de sementes: Progressos e Perspectivas de Desenvolvimento). [Private Seed Sector:
Progress and Perspectives]. Reserach Report No. 55P. Maputo, Republic of
Mozambique: Ministry of Agriculture and Rural Development.

 

 

.-—u

Meikle, G. W., Adda, C., Azor'na, K., Borgemeister, C., Degbey, P., Djomamou, B., &
Markham, H. R. (1997). The effects of maize variety on the density of
Protesphanus truncatus (Coleoptera: Bostrichidae) and Sitophilus zeamais
(Coleoptera: Curculionidae) in postharvest stores in Benin Republic. Journal of

Stored Prod. Res. 34, (1), 45-58.

Mozambique News Agency (AIM). (2007). AIM Reports No.336, 5th March 2007.
Retrieved May 10, 2007, from http://www.poptel.org.uk/mozambique-news/

Mozambique News Agency (AIM). (2007). AIM Reports No.337, 20th March 2007.
Retrieved May 10, 2007, from http://www.poptel.org.uk/mozambique-news/

Murphy, K. P., & Alexander, A. P. (2004). Persuasion as a dynamic, multidimensional
process: An investigation of individual and intraindividual differences. American
Educational Research Journal, 41, (2), 337-363.

Neupane, R. P., Sharma, K., R., & Thapa, G. B. (2002). Adoption of agroforestry in the
hills of Nepal: a logistic regression analysis. Journal of Agricultural Systems, 72,

177-196.

Nichols, P. (1991). Social Survey Methods. A F ieldguide for Development Workers.
Development Guidelines (No. 6). Oxfam.

195

Nisbett, R. E., & Ross, Lee. (1980). Theory maintenance and theory change. In Human
Inference: Strategies and shortcomings of judgment: Englewood Cliffs N. J:
Prentice Hall.

Oskamp, S., & Schultz, WP. (2005). Attitudes and Opinions (Third Edition). Mahwah,
New Jersey: Lawrence Erlbaum Associates, Publishers.

Pannell, J. D., Marshall, R. G., Barr, N., Curtis, A., Vanclay, F., & Wilkison, R. (2006).
Understanding and promoting adoption of conservation practices by rural
landholders. Forthcoming in Australian Journal of Experimental Agriculture,

PARPA. (2001). Action Plan for the Reduction of Absolute Poverty. Maputo,

Mozambique: Ministry of Planning and Finance.
Programa Nacional para Desenvolvimento Agrario (PROAGRI) [Agriculture
Development National Programme]. (2003). Maputo, Mozambique: Ministry of
Agriculture.
Rogers, E. M. (1995). Diffusion of Innovations (Fourth Edition). New York: Free i
Press.

 

Rogers, E. M., & Havens, E. (1961). The Impact of Demonstrations on Farmers’
Attitudes Toward Fertilizers: Research Bulletin No. 898. Wooster, Ohio: Ohio
Agricultural Experiment Station.

Rotter, R., & van Keulen, H. (1997). Variations in response to fertilizer application in
the tropics: II. Risks and opportunities for smallholders cultivating maize on
Kenya’s arable land. Journal of Agricultural Systems, 53, 69-95.

Ryan, B., & Gross, N. C. (1943). The diffusion of hybrid seed corn in two Iowa
communities. Rural Sociology 8, 15-24.

Rubey, L. Ward, R. W. & Tschirley, D. (1997). Maize Research Priorities. In D.
Byerlee, & C. K. Eicher (Eds), Africa ’s emerging maize revolution (pp. 145-155).

Colorado: Lynne Rienner Publishers.

Sasakawa Africa Association. (2007) Sasakawa-Global 2000 Programme. Retrieved
October 18, 2007, from http://www.saa-tokyo.org/english/aboutsaa

SEEDCO. (2003). Seed manual 2003-2004. Harare, Zimbabwe: SEEDCO, Ltd

Shaw, M. E., & Wright, J. M. (1967). Scales for the measurement of attitudes. New
York: McGraw-Hill.

SPSS 15.0 for Windows, Release 15.0.0. 2006. Chicago: SPSS Inc.

196

Suvedi, M., Lapinski, M. K., Campo, S. (2000). F armers' Perspectives of Michigan State
University Extension: Trends and Lessons from 1996 and 1999. Journal of
Extension, 38, (1).

Swanson, B. E., Roling, N. & J iggins, J. (1984). Extension Strategies for
Technology Utilization. In Swanson, B. E. (Ed.), Agricultural Extension. A
Reference Manual (second Edition). (pp. 95). Rome: FAO.

Tesfaye, Z., & Alemu, H. (2001). Adoption of Improved Maize Technologies and
Inorganic Fertilizer in Northwestern Ethiopia: Research Report No. 40. Ethiopian
Agricultural Research Organization.

Tesfaye, Z., Bedassa, T. & Shiferaw, T. (2001). Adoption of High Yielding Maize 7
Technologies in Maj or Maize Growing Regions of Ethiopia: Research Report No. '
41. Ethiopian Agricultural Research Organization.

Thompson, P. T. (1992). Perceptions and attitudes of farmers about a new fertilizer
technology. International fertilizer development center.

 

TIA (Trabalho de Inquérito Agricola) [Agricultural Survey]. 2005. Preliminary Results. #5.:
Maputo, Republic of Mozambique: Ministry of Agriculture and Rural
Development.

Uaiene, N. R. (2004). Maize and sorghum technologies and the effects of marketing
_ strategies on farmers’ income in Mozambique. Unpublished Doctoral Dissertation.
Purdue University. Indiana.

Valente, T. W. (1993). Diffusion of innovations and policy decision-making. Journal of
Communication, 43 (1), 30-45.

Visser, S. P., & Krosnick, J. A. (1998). Development strength over the life cycle: surge
and decline. Journal of personality and social psychology, 75, (6), 1389-1410.

Vowotor, K. A., Bosque-Perez, N. A., & Ayertey, J. N. (1994). Effect of Maize Variety
and Storage Form on the Development of Maize Weevil, Sitophilus zeamais
Motschulsky. Journal of stored Prod. Res. 31, (1), 29-36.

Wang, S. W., Lassoie, J. P., & Curtis, P. D. (2006). Farmer attitudes toward
conservation in J igme Singye Wangchuck National Park, Bhutan Journal of
Environmental Conservation, 33, (2), 148-156.

Willock, J ., Deary, I.J., McGregor, M., Sutherland, A., Edwards-Jones, G., Morgan, 0.,
Dent, B., Grieve, R., Gibson, G., & Austin, E. (1999). Farmers’ Attitudes,
Objectives, Behaviors, and Personality Traits: The Edinburgh Study of Decision
Making on Farmers. Journal of Vocational Behavior, 54, 5-36.

197

World Bank. (2004). Agriculture Investment Sourcebook. Agriculture and Rural
Development. Washington, D. C: The World Bank.

‘. I;
‘V
'

 

“v.1
43.

198

F v _ :11 .-
uu-In.‘--.I..-.....v....... .--.. -.u-).-. - - ' ,“-' .‘1 A “"- ‘ ‘~ ~‘ ' ‘.

u"(tullljtjjjjj(gt)l