MAU J AGR NAT (2022) 2(2): 44-55 dergipark.org.

tr/en/pub/maujan
DOI: - [email protected]
e-ISSN: 2822-3500 maujan.alparslan.edu.tr

MUŞ ALPARSLAN ÜNİVERSİTESİ MUŞ ALPARSLAN UNIVERSITY

TARIM VE DOĞA DERGİSİ JOURNAL OF AGRICULTURE AND NATURE

Adoption potentials of Urea Deep Placement (UDP) technology among smallholder rice
farmers in Kano state of Nigeria

Sanusi Mohammed Sadiq 1 • Invinder Paul Singh 2 • Muhd. Makarfi Ahmad 3 • Mahmud Sarki 4
1 Department of Agricultural Economics and Extension, FUD, Dutse, Nigeria
2 Department of Agricultural Economics, SKRAU, Bikaner, India
3 Department of Agricultural Economics and Extension, BUK, Kano, Nigeria

4 Department of Education, FUD, Dutse, Nigeria

 Corresponding Author: [email protected]

Please cite this paper as follows:

Sadiq, S. M., Singh, I. P., Ahmad, M. M., & Sarki, M. (2022). Adoption potentials of Urea Deep Placement (UDP) technology among
smallholder rice farmers in Kano state of Nigeria. Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55.

Research Article A B S T R A C T

Article History Urea deep placement (UDP) technology, a package introduced by USAID MARKETS II is
Received: 18.04.2022
challenged with limited scope in adoption and diffusion, thus the need to investigate the
Accepted: 18.05.2022
Published online: 07.06.2022 obstacles affecting its adoption potential in the studied area. A cross sectional data elicited
through a well-structured questionnaire coupled with interview schedule from a sample size
of 300 respondents: adopters (192) and non-adopters (108) sampled via multi-stage sampling
technique was used. The collected data were analyzed using Heckman’s model, Treatment
effect model and Oaxaca-Blinder decomposition model. From the empirical findings it was
established that improved seed varieties simultaneously influenced the adoption level and
Keywords:
Adoption use-intensity of UDP technology. Further, it was found that knowledge awareness on the
Nigeria technology through USAID MARKET II, promoters of the technology, had significant impact
Potential on the adoption rate of the technology. However, the potential adoption rate was hampered
Rice
due to incomplete diffusion, thus created a gap that hovers around 36.85-43.33%. Also, it was
Smallholders
UDP technology established that interaction effect on one hand- threefold decomposition and adoption
discrimination effect termed project effect on the other hand- twofold decomposition were
the major determinants of yield gap between the adopters and non-adopters. Generally, it
was concluded that the prospect of the technology is very bright in the studied area.
Consequently, the study recommends that the promoter of the project should adopt an
effective cost approach i.e. farmer-to-farmer extension approach in scaling-up the rate of
adoption of the UDP technology which exhibited a promising prospect in the studied area.

been sufficient to alleviate poverty, assure food security, or

1. INTRODUCTION
generate long-term economic growth (Kinuthia and Mabaya,

For nearly four decades, several African countries have 2017). Governments and development groups have long

been modernizing their agricultural sectors in order to ensure pushed agricultural technologies as viable means to boost

strong, long-term economic growth, food security, and farm output and decrease poverty (Ruzzante et al., 2021).

poverty reduction. Sadly, agricultural expansion has not

44
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

Increased agricultural productivity through the adoption, low agricultural output in Sub-Saharan Africa. As a result, in
diffusion of improved agricultural technologies and practices many countries, food insecurity and poverty have become
has been considered as one of the viable means of achieving the norm. Nonetheless, in order to improve the soil fertility
economic growth and agricultural transformation in low- of Africa's arable regions, particularly Nigeria, agricultural
income countries in the face of natural resource scarcity and innovations such as mineral fertilizer must be researched
climate uncertainty (Kumar et al., 2020). However, many (Donkoret al., 2019).
ostensibly beneficial technologies are nonetheless
One of the main goals of Nigerian agricultural
underutilized (Ruzzante et al., 2021). Adoption of a broad
development plans and strategies is to transition from low-
mix of such technologies and practices is often required to
productivity subsistence agriculture to a high-productivity
meet the multifaceted goals of efficiency, profitability,
agro-industrial economy. That is, moving away from old
environmental sustainability, and climate resilience (Kumar
production practices and toward new, science-based
et al., 2020).
production methods that involve new technical components
The development and dissemination of revolutionary and/or perhaps entire farming systems (Hassen, 2014).
agricultural technology is seen as a way to enhance output on
Smallholder’s adoption of new agricultural technologies
the world's 475 million small farms (less than 2 hectares),
is seen as the most important avenue out of poverty.
many of which are found in low- and middle-income
Adoption, if done effectively, should boost productivity and
countries (Lowder et al., 2016). Agriculture's responsible
offer more cash to farmers. In this sense, technology adoption
growth, according to most development experts, is a critical
can help millions of farmers escape poverty by accelerating
component in achieving, at the very least, sustainable
economic growth and creating marketing opportunities.
development goals. 1 (no poverty), 2 (zero hunger), 3
Better agricultural technology acceptance rates, on the other
(excellent health and well-being), 10 (reduced inequalities),
hand, have been disappointing and far from complete, and
12 (responsible consumption and production), 13 (climate
identifying the fundamental barriers to adoption remains a
action), 14 (life below water), and 15 (life on land) are the
challenge (Wossen et al., 2015; Wossen et al., 2017). Adoption
goals.
of new agricultural technologies is essential for the transition
Since the mid-twentieth century, the international to a more sustainable farming system, as well as a driving
community has substantially invested in the development of factor for increasing agricultural output (Obayelu et al.,
technology that increases yields, limit exposure to 2016).
environmental shocks, produce more nutritious crops,
Rice productivity should be raised to fulfill the expanding
reduces human labor requirements, and promotes long-term
population's food need, while also taking into account the
sustainability (Pardey et al., 2016a). According to evaluations
shrinking amount of land accessible for farming. This
of this investment, there was a very efficient use of public
necessitates the careful application of agricultural inputs,
funds (Pardeyet al., 2016a). Aggregate estimates of
such as high-quality seeds and fertilizers, as well as irrigation
agricultural research effectiveness; however, conceal a great
water management and other best agricultural practices
deal of heterogeneity: a few initiatives drive the average up,
(Islam et al., 2018). The most significant practices in rice
while many others fail (Pardey et al., 2016b). A high number
production are fertilizer application and water control.
of people adopting a new technology is a necessary (but not
Although nitrogen fertilizer is important in rice cultivation,
sufficient) condition for it to make an impact.
all fertilizers should be used in a balanced manner to improve
Because the vast majority of the world's poor live in rural crop output and soil fertility.
areas and work in agriculture, attempts to reduce their
The cost of nitrogen fertilizer in rice production accounts
vulnerability frequently focus on improving agricultural
for up to one-third of the entire production costs. It is
techniques as a means of increasing production, efficiency,
inefficient to apply it as granules since only approximately
and, eventually, income (Parvan, 2011; Zaidi and Munir,
one-third of the nitrogen is used up and the rest is lost.
2014). The introduction of new agricultural technology
Forming Urea into "Briquettes" and placing them deeply in
appears to bring a significant rise in output and income
the soil is one way to reduce nitrogen losses and improve
(Zaidi and Munir, 2014).
fertilizer efficiency. This IFDC (International Centre for Soil
Rural poverty and productivity shortages in developing Fertility and Agricultural Development)-developed
countries, particularly in Sub-Saharan Africa (SSA), are technique is currently widely used in Asia and is being
largely explained by the slow adoption of modern extended into Africa. Urea has risen to prominence as a key
agricultural technologies (Hörner et al., 2022). Poor soil nitrogen fertilizer for rice production, with data indicating
fertility induced by land degradation is one of the causes of

45
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

that about 80% of urea is used in rice production (Hoque et people engaged in agricultural commodities trading, making
al., 2013). it well-known for its commercial activity.

When compared to the broadcast method, deep The sample size generated from multi-stage sampling
placement of all needed fertilizers may be more efficient and technique for the study is composed of 300 rice farmers (192
farmers may benefit more. In addition, urea deep placement adopters and 108 non-adopters). For the adopter, six out of
(UDP) enhances nitrogen utilization efficiency (NUE) by up the nine project located Local Government Areas (LGAs)
to 80% (Huda et al., 2016). As a result, when compared to were purposively selected given their comparative
broadcast, UDP in rice production reduces nitrogen fertilizer advantage in cultivation of rice. The selected LGAs were
requirements by 30-35 percent while increasing grain yields Bunkure, Garun-Mallam, Kura, Dambatta, Bagwai, and
by up to 15-20 percent. In nations where nitrogen fertilizer Makoda. This was followed by random selection of five
subsidies exist, UDP minimizes government subsidy participating localities from each of the selected LGAs.
payments in addition to increasing farm profitability (Miah Thereafter, each of the chosen localities from LGAs-Bunkure,
et al., 2016; Islam et al., 2018). As a result of the better nitrogen Garun-Mallam and Kura had selection of nine (9)
"uptake" efficiency given by the bigger urea particle size and respondents each while Dambatta, Bagwai, and Makoda had
the "point placement" mode of administration, the adoption selection of four (4) respondents each, thus given a total
of UDP technology has two key benefits: increased yields and sample size of 195 adopters. For the non-adopters- control
decreased fertilizer costs. Farmers like UDP because it saves group, given the absence of a definite sampling frame, a
them money and time because they only have to fertilize rice representative sample size of 108 respondents was generated
once instead of two or three times as with the broadcast by adopting the error margin formula proposed by Bartlett et
approach and it causes fewer weeds to develop. al.(2001) (Equation 1). The distribution of the non-adopters
was done in accordance with that of the adopter category. For
Although there has been a lot of interest in the factors that
the first three LGAs, from each of the selected five (5)
influence the adoption of this technology and its practices, as
localities, five non-adopters were randomly selected; while
well as the spread of information and the impact of
from the remaining three LGAs two (2) non-adopters were
interventions that encourage them, there are still some
randomly selected from each of the chosen localities.
knowledge gaps in the field. As a result, it's critical that this
Objectives I, II and III were achieved using Heckman’s
new information on UDP technology's potential adoption
model, Treatment estimation model and Oaxaca-Blinder
rate be leveraged to drive more usage. Improving access to
decomposition model respectively. A well structured
this technology not only benefits smallholder farmers, but it
questionnaire coupled with interview schedule was used to
can also benefit the rural poor because increased production
elicit information using easy-cost route approach owing to
leads to lower rice prices. Consequently, this research aimed
farmers’ memory recall rather than book keeping which is
at determining the potential adoption rate of UDP
absent among the respondents. It is worth to note that three
technology in the study area keeping in view the specific
questionnaires of the adopters had incomplete information,
objectives viz. determinants of adoption level and intensity
as such eliminated, thus living a total of 192.
of technology use; adoption potential rate of the technology;
 P( P  1)
2
and, the effect of adoption discrimination on yield.
nn  Z  (1)
2
2. MATERIAL AND METHODS e
Where, n= finite sample size; Z = 1.645 (t-statistic at 10%
Kano state is in northern Nigeria, with latitudes ranging
level); P = 80% (proportion); and, e = error gap at 10% degree
from 10° 33 to 12° 37N and longitudes ranging from 07° 34 to
of freedom (0.10).
09° 25E of the Greenwich meridian time. The northern and
southern portions of the state's vegetation are characterized Heckman’s Model: The model is made up of two parts: a
by Northern-Guinea savannah and Sudan savannah, choice model and an outcome model, with the former having
respectively. The yearly rainfall in the Northern-Guinea a dichotomous dependent component and the latter having a
savannah ranges from 600-1200 mm to 300-600 mm in the continuous predict variable (Sadiq et al., 2021a).Because of its
Sudan savannah. Furthermore, in the Sudan savannah capacity to correct sample selection bias, the two-step
region, arable crop growth seasons range from 90 to 150 days, Heckman's selection model was chosen. The model is shown
and in the Northern-Guinea savannah region, they range below:
from 150 to 200 days. The population of the state is predicted
𝑌𝑖 = 𝑓(𝑋1 , 𝑋2 , 𝑋3 … … … … … … . . 𝑋𝑛 )…….…………………. (2)
to reach 9.4 million people by 2050 (NPC, 2006), with a 3.5
percent annual growth rate.There are around 1,754,200 𝑌𝑖𝑡 = 𝛽0 + 𝛽𝑋𝑖𝑡 + 𝜀𝑖 ……….……………………………….… (3)
hectares of arable land in the state. The bulk of the state's
46
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

𝑌𝑖∗ = 𝛼 + 𝑋𝛽 + 𝜀𝑖 ....................................................................... (4) 𝐴𝑇𝑈 = 𝐸( 𝑦1𝑖 |𝐼 = 1; 𝑋) − 𝐸( 𝑦2𝑖 |𝐼 = 1; 𝑋) … … … … … … … . (11)

𝑌𝑖∗ = 𝛼 + 𝑋1 𝛽1 + 𝑋2 𝛽2 + 𝑋3 𝛽3 + 𝑋4 𝛽4 + 𝑋5 𝛽5 + ⋯ . +𝑋𝑛 𝛽𝑛 + Average Treatment Effect on Treated = ATET

𝛾𝐼𝑀𝑅 + 𝜀𝑖 ……………………………………………………… (5) Average Treatment effect on Untreated = ATEU

Equation 3 is a decision stage, a probit model with the Equations (10) and (11) were further simplified as:
dependent variable been binary while Equation 5 is an
𝑁1
outcome stage, a censored model with the dependent 1
variable been continuous. ATT = ∑[𝑝 (𝑦1𝑖 |𝐼 = 1; 𝑋) − 𝑝( 𝑦2𝑖 | 𝐼 = 1; 𝑋)] … . … … (12)
𝑁1
𝑖=1
Where, 𝑌𝑖𝑡 = 𝐴𝑑𝑜𝑝𝑡𝑖𝑜𝑛 𝑐𝑎𝑡𝑒𝑔𝑜𝑟𝑦(𝑎𝑑𝑜𝑝𝑡 = 1, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 =
𝑁2
0); Yi* = latent observation of ith adopter (proportion (%) of 1
ATU = ∑[ 𝑝 ( 𝑦2𝑖 |𝐼 = 0; 𝑋) − 𝑝( 𝑦1𝑖 | 𝐼 = 0; 𝑋)] … … . (13)
farm size under UDP technology); 𝑋1 = Gender (male = 1, 𝑁2
𝑖=1
otherwise=0); 𝑋2 = Age (year); 𝑋3 = Marital status (married =1,
otherwise=0); 𝑋4 = Educational level (year); 𝑋5 = Household Where, 𝑁1 and 𝑁2 are number of aware and non-aware
size (number); 𝑋6 = Rice farming experience/ farming farmers respectively, and 𝑝= probability.
experience (year); 𝑋7 = Mixed cropping (yes= 1,
otherwise=0); 𝑋8 = Extension contact (yes=1, otherwise=0); 𝑋9 Oaxaca-Blinder Decomposition Model
= Seed variety (improved =1, otherwise=0); 𝑋10 = Duration of
UDP adoption (year); 𝑋11 = Total farm size/ rice farm size The following is the yield index function (Oaxaca 1973;
(ha); 𝑋12 =Co-operative membership (yes=1, otherwise=0); 𝑋13 Blinder 1973; Sadiq et al., 2020):
= Total livestock unit (TLU) (Camel=1.0; Horse=0.8;
𝑖
Cattle=0.7; Donkey=0.5; Sheep & Goat =0.1; and,
𝑌̅𝐴 = 𝛽0 + 𝛽𝑖 ∑ 𝑋𝑖 + 𝜀𝑖 … … … … … … … … … … … … … … … . . (14)
Chicken=0.01); 𝑋14 = Market distance (km); 𝑋15 = 𝑖=1
Commercialization index (CI)(ratio of marketed surplus to
𝑖
marketable surplus); 𝑋16 = Dead stocks (capital assets);𝑋17 =
𝑌̅𝑁𝐴 = 𝛽0 + 𝛽𝑖 ∑ 𝑋𝑖 + 𝜀𝑖 … … … … … … … … … … … … … … … . (15)
Yield (kg); IMR= The inverse Mill’s ratio;𝛽0 = Intercept;𝛽1−17
𝑖=1
= Regression coefficient; 𝛾= Lambda; and, 𝜀𝑡 = Stochastic.
Where, 𝑌̅𝐴 = average yield of adopters; 𝑌̅𝑁𝐴 = average yield
Average Treatment Effect (ATE)
of non-adopters; 𝑋𝑖−𝑛 = 𝑒𝑥𝑝𝑙𝑎𝑛𝑎𝑡𝑜𝑟𝑦 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠; 𝛽0 =
The average result difference between units assigned to 𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡; 𝛽𝑖−𝑛 = 𝑝𝑎𝑟𝑎𝑚𝑒𝑡𝑒𝑟 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒𝑠; and, 𝜀𝑖 =
care and units assigned to placebo is depicted in this graph 𝑠𝑡𝑜𝑐ℎ𝑎𝑠𝑡𝑖𝑐 𝑡𝑒𝑟𝑚.
(control). The following equation is based on Sadiq et
The total difference can be explain by,
al.(2021b):

∆𝑌 = 𝑌̅𝐴 − 𝑌̅𝑁𝐴 ……………………………………………... (16)

Awareness index is given by: 𝐸(𝑦1𝑖 |𝐼 = 1; 𝑋) … … … … … … . (6)

The Oaxaca-Blinder decomposition equation is,

Non-Awareness index is given by: 𝐸( 𝑦2𝑖 |𝐼 = 0; 𝑋) … … … . (7)

∆𝑌 = (𝑋̅𝐴 𝛽̂𝐴 − 𝑋̅𝑁𝐴 𝛽̂𝑁𝐴 ) + (𝑋̅𝑁𝐴 𝛽̂𝐴 − 𝑋̅𝐴 𝛽̂𝑁𝐴 ) …………… (17)
Index of the awareness if there is no difference is denoted by:
𝐸( 𝑦2𝑖 |𝐼 = 1; 𝑋) … … … … … … … … … … … … … … … … … … … … (8)
If there is only discrimination against the non-adopters
Index of non-awareness if there is difference is denoted the formula becomes:
by:𝐸( 𝑦1𝑖 |𝐼 = 0; 𝑋) … … … … … … … … … … … … … … … … . … … (9)
∆𝑌 = (𝑋̅𝐴 − 𝑋̅𝑁𝐴 )𝛽̂𝐴 + (𝛽̂𝐴 − 𝛽̂𝑁𝐴 )𝑋̅𝑁𝐴 + (𝑋̅𝐴 − 𝑋̅𝑁𝐴 )(𝛽̂𝐴 − 𝛽̂𝑁𝐴 )
Where: ……………………………………………………………… (18)
𝐸(. ) = Expectation operator
If the non-adopter group has adopter group’s coefficient,
𝑦1𝑖 = UDP Adoption (dependent variable)
then the formula becomes:
𝑦2𝑖 = Non-Adoption (dependent variable)
𝐼 = Dummy variable (1 = awareness, 0 = non-awareness
∆𝑌 = (𝑋̅𝐴 − 𝑋̅𝑁𝐴 )𝛽̂𝑁𝐴 + (𝛽̂𝐴 − 𝛽̂𝑁𝐴 )𝑋̅𝑁𝐴 + (𝑋̅𝐴 − 𝑋̅𝑁𝐴 )(𝛽̂𝐴 −
𝑋 = Explanatory variables that is common to both.
𝛽̂𝑁𝐴 )………………………………………………………… (19)

ATT = 𝐸( 𝑦1𝑖 |𝐼 = 1; 𝑋) − 𝐸( 𝑦2𝑖 |𝐼 = 1; 𝑋) … … … … … … . . (10)

47
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

Thus, Equations (18) & (19) have a ‘threefold’ technology- a precursor and catalyst, the major farm’s goals
decomposition, i.e. the outcome difference is divided into will be in jeopardy. Therefore, the tendency of farmers that
three components. The first, second and third components used improved seed varieties to adopt UDP over their
respectively, are endowment effect, discrimination effect and counterparts (users of local varieties) will be 50.82% while the
interaction effect. intensity of the technology usage will increase by 3.24% vis-
à-vis 8.66 hectares. This high acreage increase is mostly likely
The idea that there is a non-discriminatory coefficient among the medium to large-scale farmers if they observe
vector that should be utilized to determine the contribution economies of scale given their access to agricultural holdings.
of the variations in the predictors leads to an alternative
The negative-significant of rice farm size implies that
decomposition that is popular in the discrimination research.
diseconomies of scale plummets the intensity of UDP
Let 𝛽 ∗ be a non-discriminatory coefficient vector of this type.
adoption among the medium to large scale farmers. If the
Following Jan (2008), the outcome difference can then be
investment returns is not commiserate viz. increase in cost of
written as:
cultivation vis-à-vis increase in yield, which in turns hampers
∆𝑌 = (𝑋̅𝐴 − 𝑋̅𝑁𝐴 )𝛽∗ + {(𝑋̅𝐴 )(𝛽̂𝐴 − 𝛽∗) + (𝑋̅𝑁𝐴 )(𝛽∗ − 𝛽̂𝑁𝐴 )} the enterprise going concern due to mismatch in cost-benefit
……………………………………………………………… (20) ratio, there is the tendency among the large-scale farmers to
scale-down the intensity of UDP usage in the study area.
Therefore, Equation (20) has a ‘twofold’ decomposition, Therefore, the marginal and elasticity implications of large-
i.e. the outcome difference is divided into two components. scale farmers scaling down their intensity in the use of UDP
The first and second are quantity effect and unexplained technology viz. cultivation size due to diseconomies of scale
effect respectively. The latter is frequently attributed to is likely to be 6.12 hectares and 7.65% respectively.
discrimination, but it's vital to remember that it also includes The negative-significant of gender coefficient indicates
all of the possible effects of unobserved variables. that poor access to productive resources among the women
farmers affect them in the adoption of UDP technology.
3. RESULTS AND DISCUSSION Gender stereotype due to religion and culture in the study
area remains a challenge that hinders women’s active
Determinants of UDP’s adoption level and use intensity
participation in the downstream sector of agriculture, thus
The Wald Chi2 being different from zero at 10% degree of triggering their vulnerability to the vicious cycle of poverty.
freedom alongside the absence of interdependence- Therefore, the likelihood of women farmers not adopting this
multicolinearity between the predictors as evident by their technology against their men counterparts is 67.26%. The
respective variance inflation factors (VIF) that are within the coefficient of marital status being negatively signed and
threshold value of 10.0 means that the chosen model is best significant reveals that poor access to capital viz. social and
fit for the specified equation (Table 1). Further, the non- economic capitals alongside less family responsibility to cater
plausibility of the Lambda’s estimated coefficient viz. inverse for affects the adoption of UDP technology among the
Mill’s ratio at 10% reinforced the validity, reliability and farmers that are single. The social and economic capitals
accuracy of the chosen model for future prediction as it which are inherent in marriage are cogent veritable tools that
implies that the model has no problem of sample selection will enable married farmers to be able to afford the resources
bias in the use of non-zeros adoption intensity values. to adopt the technology. In addition, the need to have an

A perusal of the results showed that improved seed enlarged income so as to have access to better living standard

varieties influenced both adoption level and intensity of UDP for farm family will motivate married farmers to adopt the

technology usage; gender, marital status, household size, technology. Thus, the probability of non-married farmers not

extension contact and distance to market had influence on adopting the UDP technology in comparison to their

adoption status; while, rice farm size influenced adoption counterparts (married farmers) is 73.93%.

intensity as indicated by their respective parameter estimates The empirical evidence shows that large farm families are
that are within the acceptable margin of 10% error gap. less likely to adopt the UDP technology as evident by the

The positive-significant of improved seed varieties imply household size coefficient that is negatively signed and

that UDP’s adoption and intensity of use is high among the significant. The possible reason for poor adoption of the

farmers that used improved seed varieties when compared technology among this category of farmers may be attributed

to their counterparts that used non-improved (local) to population pressure that owes to the households being

varieties. The possible reason may owes to the need to have composed of vulnerable people- women and children: non-

potential yield level, which without the use of the able bodied people, with little or no impact on external
remittances. Thus, the probability of large household size
48
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

opting out of UDP technology adoption will be 3.2% for any Therefore, the probability of farmers with extension access
increase in household composition by one person as shown adopting the technology against their counterparts without
by the elasticity estimate of the respective variable. is 108%. Closeness to the source of UDP technology
encourages the adoption of the technology as indicated by
The results showed that adoption of the technology was
the positive-significant of the estimated coefficient associated
high among farmers that have access to extension services
with distance to market. Proximity to markets where there
against their counterparts with no access as indicated by the
exist perfect market competition vis-à-vis the input cost and
extension estimated coefficient that is significant and
readily available instructional information on the use of the
positively signed. Proper guidance and support services
technology among various sorts of agro-allied service dealers
from advisory services are impetus for attainment of
play a crucial role in motivating farmers at close distance to
potential farm objectives, thus strong motivational catalysts
the market to adopt the technology. Therefore, the possibility
that encourages adoption of technologies especially among
of farmers with close proximity to markets to adopt the UDP
the mammoth/ large group of low educated farmers that
technology is 16.20% against their counterparts whose
characterized the farming settings in the study area.
proximity to the markets are far.

Table 1a. Determinants of adoption and use intensity

Adoption status (decision) Adoption intensity (outcome)
Items Elasticity
Coefficient t-stat Coefficient t-stat
Intercept −1.414(0.876) 1.613ns 72.36(32.65) 2.216** -
Gender −0.672(0.306) 2.197** - - -
Age 0.005(0.010) 0.537ns - - -
Marital status −0.739(0.323) 2.285** - - -
Education 0.011(0.018) 0.645ns 0.247(0.325) 0.760ns 0.0231
Household size −0.032(0.017) 1.863* - - -
Experience 0.019(0.013) 1.437ns - - -
Experience (rice) - - 0.209(0.232) 0.901ns 0.0416
Mixed cropping −0.503(0.308) 1.633ns - - -
Extension contact 1.083(0.371) 2.914*** −18.04(14.15) 1.274ns -0.2643
Seed variety 0.508(0.209) 2.429** 8.663(4.729) 1.832* 0.0324
DUDPA - - 0.292(0.903) 0.324ns 0.0108
Farm size (total) −0.020(0.065) 0.308ns - - -
Farm size (rice) - - −6.123(3.467) 1.766* -0.0765
CM 0.135(0.267) 0.504ns - - -
TLU 0.055(0.060) 0.919ns - - -
MD 0.162(0.047) 3.409*** - - -
CI 0.547(0.455) 1.201ns - - -
lnDS 0.099(0.072) 1.364ns - - -
Yield - - 0.803(3.314) 0.242ns 0.0987
Lambda(IMR) - - 1.009(9.767) 0.103ns -
Sigma 24.602
Rho 0.0410
Wald Chi2 10.19[0.0177]**
Source: Field survey, 2018
Note: *** ** * &ns imply significant at 1%, 5%, 10% & non-significant, respectively.
DUDPA= Duration of UDP adoption; CM= Co-operative membership; MD=Market distance; lnDS= log of Dead stock.
Figures in ( ) and [ ] are standard error and probability level, respectively

Table 1b. Multicollinearity test significant impact on the adoption rate of UDP technology
Variables Variance inflation factors in the studied area as evident by their respective ATE
Education 1.076 estimated coefficients which are positively signed and within
Experience (rice) 1.144
Extension contact 1.064
the plausibility of 10% probability level (Table 2). The ATE
Seed variety 1.065 coefficients of propensity score matching (PSM), regression
DUDPA 1.107 adjustment (RA), nearest-neighbor matching (NNM), inverse
Farm size (rice) 1.288
probability weights (IPW) and IPW regression adjustment
Yield 1.168
(IPWRA) being 0.1967, 0.2615, 0.2083, 0.2311 and 0.2287
UDP’s Adoption Potentials respectively, implies that the awareness via USAID
MARKETS II increase the adoption rate of UDP technology
A perusal of the ATE results vis-à-vis all the estimation
by 19.67, 26.15, 20.83, 23.11 and 22.87% respectively.
techniques shows that knowledge gained from participation
in the Urea displacement project (UDP) had a positive-
49
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

Furthermore, within the participating group, the knowledge awareness impacted significantly on the
empirical evidence establish that diffusion had positive adoption and diffusion of the technology in the study area.
impact on the adoption potential of the technology as
indicated by all the respective estimation techniques of ATET Further, based on the cursory review of the impact of the
coefficients which are positive signed and within the knowledge on the use of the UDP technology, the empirical
acceptable margin of 10% probability level. The ATET evidences of the diffusion of the technology showed that only
coefficients of the PSM, RA, NNM, IPW and IPWRA being 63% of the farming population was aware of the technology
0.1587, 0.2306, 0.1958, 0.1941 and 0.1907 respectively, means (Table 2). This incomplete diffusion of the technology limits
that diffusion due to programme participation increase the the adoption rate to 64% when the potential adoption rate is
potential adoption rate of the UDP technology among the 98.44%, thus leads to an adoption gaps that hovers around
adopters in the study area by 15.87, 23.06, 19.58, 19.41 and 36.85 to 43.33%. Given the estimates of PSM, RA, NNM, IPW
19.07% respectively. However, on the other side, poor and IPWRA, the adoption gaps are 43.33, 36.85, 42.17, 39.89
diffusion of the technology negatively affected the adoption and 40.13% respectively. Based on the selection bias vis-à-vis
potential rate of the technology among the non-participating most of the estimation techniques, it can be inferred that all
group as indicated by the ATEU of all the respective the farmers have an equal opportunity to adopt this
estimation techniques which are negatively signed and technology. This demonstrates the consistency of the
different from zero at 10% probability level. The ATEU adoption of UDP technology among all the farmers in the
estimated coefficients of the PSM, RA, NNM, IPW and studied area. Besides, it can be concluded that the UDP
IPWRA being -0.2613, -0.3140, -0.2297, -0.2902 and -0.2779 technology has a promising prospect in the studied area. In
respectively, implies that poor diffusion rate among the non- examining the UDP adoption rate across the estimated
participating group plummeted the adoption potential of techniques, the PSM, RA, NNM, IPW and IPWRA showed
UDP technology by 26.13, 31.40, 22.97, 29.02 and 27.79% 63% of the adopters with the possibility of gaining 19.67,
respectively. Therefore, it can be inferred that the programme 26.15, 20.83, 23.11 and 22.87% potential adoption rates
respectively.

Table 2. Adoption potentials of UDP technology

ATE ATET ATEU
Items Coeff. SE t-stat Coeff. SE t-stat Coeff. SE t-stat
PSM 0.196667 0.07755 2.54*** 0.15873 0.066989 2.37** -0.26126 0.146637 1.78*
RA 0.261483 0.061443 4.26*** 0.230617 0.065277 3.53*** -0.31404 0.076917 4.08***
NNM 0.208333 0.07044 2.96*** 0.195767 0.072242 2.71*** -0.22973 0.112062 2.05**
IPW 0.231141 0.059486 3.89*** 0.19409 0.06225 3.12*** -0.29023 0.07743 3.75***
IPWR 0.228714 0.057883 3.95*** 0.190698 0.059612 3.20*** -0.27785 0.067021 4.15***
Table 2. Continued
Pop. (N) Adopt (NA) Exposed (NE) A B AG PBS A% B% C
PSM 300 192 189 0.64 0.63 -0.43333 -0.03794 64 63 98.4375
RA 300 192 189 0.64 0.63 -0.36852 -0.03087 64 63 98.4375
NNM 300 192 189 0.64 0.63 -0.42167 -0.01257 64 63 98.4375
IPW 300 192 189 0.64 0.63 -0.39886 -0.03705 64 63 98.4375
IPWR 300 192 189 0.64 0.63 -0.40129 -0.03802 64 63 98.4375

Source: Field survey, 2018

Note: Pop. = Population; A = common adoption rate (NA/N); B= common adoption and expose rate (NE/N); A% = Adoption rate
in the general population; B% = Rates of those who know; C= Rates of those who know and have adopted(NA/NE); AG = Adoption
Gap (ATE-B); PBS = Population bias selection (ATE-ATET)

Discriminate effect of UDP technology on yield level

A cursory review of the results showed the average rice differences in human factors accounted for -2.20 quintals
yield levels of the adopters and non-adopters to be 35.96 and while interaction effect holds on to 2.88 quintals. Thus, it can
34.28 quintals respectively, leaving a gap of 1.69 quintals to be inferred that interaction effect vis-à-vis combination of
be explained by the Blinder-Oaxaca Decomposition (Table 3). human factor-related and structural effects accounted for the
Of the gap, the threefold result showed discrimination effect major gap in the yield level between the adopters and non-
termed structural difference due to adoption of the UDP adopters of UDP technology in the studied area.
accounted for 1.01 quintals; endowment effect due to Furthermore, the results of Figure 1 showed most of the

50
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

variables in the endowment component to have a statistically accounted for some portion of the high yield achieved by the
insignificant influence, except distance to market. It is adopters of UDP technology. In the coefficient component,
obvious that the adopter versus non-adopter yield gap is except rice farming experience, most of the variables are
driven by group differences in the proportion of individuals’ either insignificant or exhibits marginal significant influence.
distance to market. Thus, individuals with close proximity to Thus, differences in the parameter estimates of rice farming
market tend to gain more yield as evident from the pooled experience account for decisive portion of the yield gap.
regression coefficient of distance to market. More so, the Based on the rice farming experience coefficient differences
value of x.mean.diff indicated that a large proportion of the between the two groups, it can be inferred that the yield gain
adopters of the technology are close to the market. Therefore, for an additional year of rice farming experience is higher for
it can be inferred that the difference in the distance to market the adopter group by 1.25%.
composition of the adopter and non-adopter groups

Table 3a. Summary of variables in the Oaxaca-Blinder decomposition model

Items Beta A Beta B Beta Diff. Reg. A Reg. B Reg. Pool 1 Reg. Pool2 Mean A Mean B Diff.
Intercept 39.22418 27.9987 11.22548 39.22 28 39.53 41.35 1 1 0
Gender 4.473236 13.61929 -9.14606 4.473 13.62 4.801 5.107 0.640625 0.814815 -0.17419
Age -0.19379 -0.00483 -0.18896 -0.1938 -0.00483 -0.1805 -0.1826 39.96875 39.26852 0.700232
MS 0.121721 13.70013 -13.5784 0.1217 13.7 3.315 3.73 0.885417 0.907407 -0.02199
Education -0.13209 0.63325 -0.76534 -0.1321 0.6332 0.1875 0.1772 5.989583 6.12037 -0.13079
EXR 0.464739 -0.78605 1.250789 0.4647 -0.7861 0.2142 0.2049 12.80208 12.5 0.302083
MC -9.05241 -1.57052 -7.48189 -9.052 -1.571 -5.46 -5.288 0.854167 0.916667 -0.0625
Extension 3.710732 -0.98711 4.697844 3.711 -0.9871 0.8746 0.1512 0.979167 0.814815 0.164352
DM 1.081684 2.767762 -1.68608 1.082 2.768 1.708 1.593 2.828125 1.574074 1.254051
SV 1.401076 3.52578 -2.1247 1.401 3.526 0.7851 0.4435 0.322917 0.12963 0.193287
Farm size -8.56346 -7.97406 -0.5894 -8.563 -7.974 -8.081 -8.224 0.803438 0.67213 0.131308
CM -3.25564 -10.466 7.210352 -3.256 -10.47 -5.819 -5.975 0.869792 0.675926 0.193866
TLU 2.698875 -2.73447 5.433343 2.699 -2.734 0.9675 0.9441 1.238958 1.042315 0.196644
CI 2.831055 3.786765 -0.95571 2.831 3.787 3.23 2.904 0.713387 0.670659 0.042727
DS 1.51E-05 -9.1E-06 2.42E-05 1.51E-05 -9.1E-06 -9.7E-07 -7.6E-07 85522.03 102931.6 -17409.5
Class/mean -2.223 35.96439 34.27876 1.685632

Source: Field survey, 2018

Note: Diff. = Difference; Reg. = Regression; MS = marital status; EXR = Experience (rice); MC = mixed cropping; DM = Distance to
market; SV = Seed variety; CM = Co-operative membership; DS = Dead stock

Table 3b. Gap due to discrimination of adoption (Threefold decomposition)

Endowment effect Coefficient effect Interaction effect
Items
Coeff. SE t-stat Coeff. SE t-stat Coeff. SE t-stat
Intercept 0.000 0.000 0.000ns 11.22548 20.69276 0.542ns 0.000 0.000 0.000ns
Gender -2.372 1.298664 1.826* -7.45234 7.485321 0.995ns 1.59315 1.703522 0.935ns
Age -0.00338 0.427615 0.007ns -7.42028 2.393456 3.100*** -0.13232 0.61525 0.215ns
MS -0.30128 0.073839 4.080*** -12.3211 7.036827 1.750* 0.298599 0.124248 2.403***
Education -0.08282 0.472753 0.175ns -4.68414 5.151241 0.909ns 0.100096 0.576317 0.173ns
EXR -0.23745 0.597481 0.397ns 15.63486 5.544397 2.819*** 0.377843 0.122217 3.091***
MC 0.098157 0.623244 0.157ns -6.8584 3.628618 1.890* 0.467618 0.208279 2.245**
Extension -0.16223 1.574623 0.103ns 3.827873 7.914037 0.483ns 0.772099 1.61536 0.477ns
DM 3.470914 1.961137 1.769* -2.65401 1.24442 2.132** -2.11443 1.194978 1.769*
SV 0.681488 0.310328 2.196** -0.27542 0.076215 3.613*** -0.41068 0.197192 2.082**
Farm size -1.04706 0.371833 2.815*** -0.39615 3.600049 0.110ns -0.07739 1.056069 0.073ns
CM -2.029 1.124915 1.803* 4.873664 2.378449 2.049** 1.39784 1.802532 0.775ns
TLU -0.53772 0.165448 3.250*** 5.663254 1.890864 2.995*** 1.068432 1.557341 0.686ns
CI 0.161798 0.079776 2.028** -0.64096 10.59402 0.060ns -0.04083 0.406919 0.100ns
DS 0.158054 0.992811 0.159ns 2.486803 1.887656 1.317ns -0.42061 0.20577 2.044**
Effect -2.20287 0.566421 3.889*** 1.009082 0.395203 2.553*** 2.879417 1.443967 1.994**
WD 35.28784 35.28784
% of Disc. 2.94
Source: Field survey, 2018
Note: Coeff. = Coefficient; SE = Standard error; WD = Without discrimination; Disc. = Discrimination
51
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

Table 3c. Gap due to discrimination of adoption (Twofold decomposition)

Items Explained Unexplained
Weight Coeff. SE t-stat Coeff. SE t-stat
0 -2.20287 2.566421 0.858ns 3.888499 2.723143 1.427ns
1 0.67655 0.338812 1.996** 1.009082 3.095203 0.326ns
0.5 -0.76316 0.316809 2.408** 2.44879 1.34477 1.820**
0.64 -0.36004 1.821734 0.197ns 2.045672 2.51621 0.812ns
-1 -0.06154 0.030757 2.001** 1.747168 1.024824 1.704*
-2 -0.53691 0.254261 2.111** 2.222546 1.264236 1.758*
WD(-1) 36.02593^ 36.02593^^
WD(-2) 36.5013^ 36.5013^^
% of Disc. (-1) 5.10
% of Disc. (-2) 6.48

Unexplained A Unexplained B
Coeff. SE t-stat Coeff. SE t-stat
0 3.88849 2.72314 1.427ns 0.000 0.000 0.000ns
1 0.000 0.000 0.000ns 1.009082 3.095203 0.326ns
0.5 1.94424 1.16157 1.673ns 0.504541 1.547601 0.326ns
0.64 1.39986 1.74281 0.803ns 0.645812 1.114273 0.579ns
-1 0.628981 0.26533 2.370** 1.118188 0.362207 3.087***
-2 -3.1E-14 1.32E-14 2.331** 2.222546 0.764236 2.908***

Source: Field survey, 2018

Note: A = Adoption group; B = Non-Adoption group; WD = without discrimination; ^ = mean yield of Adopters minus endowment
coefficient; ^^ = mean yield of non-adopters plus discrimination coefficient
done by Hlavac (2018); Neumark (1988). In the weight
column, the Neumark’s decomposition is denoted by (-1). A
perusal of the overall twofold decomposition results showed
that the yield gap of 1.69 quintals between the adopters and
non-adopters can be decomposed into -0.062 quintal that is
explained by the group differences in the predictor variables
and 1.747 quintals that is unexplained. While at weight (-2),
the yield gap can be decomposed into -0.537 and 2.223
quintals respectively, which owes to explained and
unexplained effects respectively. If it is assumed that the
unexplained component of the yield gap happens due to
technology discrimination, and the pooled regression
coefficients are non-discriminatory, then the results at
Figure 1. Threefold decomposition weights (-1) and (-2) revealed that 0.629 and -3.086E-14
Note: Age (A); Marital status (MS); Rice farming experience quintals respectively, of the unexplained part originates from
(REX); Mixed cropping (MC); Extension contact (EXT); the discrimination in favour of the adopters (unexplained A)
Market distance (DIS); Rice farm size (FS); Co-operative while 1.118 and 2.223 quintals in respect of the weight
membership (CP); Tropical livestock unit (TLU); Dead stock aforementioned emanates from discrimination against the
(DS) non-adopters (unexplained B). The results of the variable-by-

The results of the twofold decomposition showed two variable twofold decomposition depicted in Figure 2 are

negative weights that implies that the reference coefficients consistent with that of the threefold decomposition.

come from the pooled regression either without (-1) or with Similarly, it was observed that the yield gap is govern by the

(-2) the group indicator variable included as a covariate higher proportion of individuals with close proximity to

(Table 3). The use of the pooled regression coefficients as a market among the adopters (explained component) and

reference coefficient set excludes the group indicator likewise their greater yield gain that owes to rice farming

variables of non-adopters. This is in line with previous works experience (unexplained component).

52
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

level should be 36.03 and 36.50 quintals (approximately 36

quintals across) respectively.

Figure 2. Twofold decomposition

Note: Gender (G); education (ED); Seed variety (SV); Figure 4. Side-by-side variable comparison
Commercialization index (CI) Note: Favor of adopters- first row; against non-adopters-

In Figure 3, nine variables were used to visualize how second row

much of the unexplained portion of the yield gap can be

4. CONCLUSION
attributed to discrimination in favor of the adopters and how
much owes to discrimination against the non-adopters. From
From the findings it was inferred that improved seed
the graph, it appears that only the discrimination component
varieties has a dual influence on adoption and intensity of
of rice farming experience achieved non-marginal statistical
use of the UDP’s technology. Further, it was found that the
significance. From the bars, the relative size suggests that if it
project has positive significant influence on the adoption rate
is assumed that the pooled regression coefficients reflect a
of the technology. Besides, the potential rate of adoption was
non-discriminatory- almost thrice as much of the yield gap is
high but incomplete diffusion of the technology was found to
explained by discrimination against the non-adopters
be a barrier, thus created an adoption gap of 36.85-43.33%.
compared to the discrimination in favor of the adopters.
More so, the empirical evidences showed that the
Besides, the side-by-side variable comparison is shown in
composition of endowment-related factors and adoption
Figure 4.
discrimination effects- an interaction effect on one hand-
threefold decomposition; and, adoption discrimination effect
termed structural effect, a de facto on other hand- twofold
decomposition played key roles in determining the yield gap
between the adopters and non-adopters. Generally, it was
concluded that the prospect of the technology in the studied
area is very promising. Therefore, it was recommended that
the project should scale-up the adoption rate of the
technology preferably through farmer-to-farmer extension
approach, an effective diffusion medium for a wider
acceptability of the technology.

Compliance with Ethical Standards

Author Contributions

Authors contributed equally to this paper.

Conflict of Interest

Figure 3. Variables vis-à-vis adopters and non-adopters The authors do not have any conflicts of interest to
declare.
Further, at twofold decomposition, the empirical
evidences showed that at weights (-1) and (-2), without Ethical Approval
discrimination against the non-adopters, their average yield For this type of study, formal consent is not required.

53
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

REFERENCES Kumar, A., Takeshima, H., Thapa, G., Adhikari, N., Saroj, S.,
Karkee, M., & Joshi, P. K. (2020). Adoption and
Bartlett, J. E., Kotrlik, J. W., & Higgins, C. C. (2001). diffusion of improved technologies and production
Organizational research: determining appropriate practices in agriculture: Insights from a donor-led
sample size in survey research. Information Technology, intervention in Nepal. Land Use Policy, 95, 104621.
Learning and Performance Journal, 19(1), 43-50. https://doi.org/10.1016/j.landusepol.2020.104621
Blinder, A. S. (1973).Wage discrimination: reduced form and Lowder, S. K., Skoet, J., & Raney, T. (2016). The number, size,
structural estimates. Journal of Human Resources, 8(4), and distribution of farms, smallholder farms, and
436-455. family farms worldwide. World Development, 87, 16-
Donkor, E., Onakuse, S., Bogue, J. & De Los Rios-Carmenado, 29. https://doi.org/10.1016/j.worlddev.2015.10.041
I. (2019). Fertiliser adoption and sustainable rural Miah, Md. A. M., Gaihre, Y. K., Hunter, G., Singh, U., &
livelihood improvement in Nigeria. Land Use Policy, Hossain, S. A. (2016). Fertilizer deep placement
88, 104193. increases rice production: evidence from farmers’
https://doi.org/10.1016/j.landusepol.2019.104193 fields in southern Bangladesh. Agronomy Journal,
Hassen, B. (2014). Factors affecting the adoption and intensity 108(2), 1-8. https://doi.org/10.2134/agronj2015.0170
of use of improved forages in North East Highlands Neumark, D. (1988). Employers' discriminatory behavior and
of Ethiopia. American Journal of Experimental the estimation of wage discrimination. Journal of
Agriculture, 4(1), 12-27. Human Resources, 23(3), 279-295.
Hlavac, M. (2018). Oaxaca: Blinder-Oaxaca Decomposition in https://doi.org/10.2307/145830
R. R Package Version 0.1.4. Oaxaca, R. (1973). Male-female wage differentials in urban
Hoque, M. A., Wohab, M. A., Hossain, M. A., Saha, K. K., & labor markets. International Economic Review, 14(3),
Hassan, M. S.(2013). Improvement and evaluation of 693-709. https://doi.org/10.2307/2525981
Bari USG applicator. Agricultural Engineering Obayelu, A. E., Okuneye, P. A., Shittu, A. M., Afolam, C. A.,
International: CIGR Journal, 15(2), 87-94. & Dipeolu, A. O. (2016). Determinants and the
Hörner, D., Bouguen, A., Frölich, M. & Wollni, M. (2022). perceived effects of adoption of selected improved
Knowledge and adoption of complex agricultural food crop technologies by smallholder farmers along
technologies: evidence from an extension experiment. the value chain in Nigeria. Journal of Agriculture and
The World Bank Economic Review, 36(1):68-90. Environment for International Development, 110(1), 155-
https://doi.org/10.1093/wber/lhab025 172. https://doi.org/10.12895/jaeid.20161.436
Huda, A., Gaihre, Y. K., Islam, M. R., Singh, U., Islam Pardey, P. G., Chan-Kang, C., Dehmer, S. P., & Beddow, J. M.
Md. R., Sanabria, J. S., Satter, M. A., Afroz, H., (2016a). Agricultural R&D is on the move. Nature, 537,
Halder, A., & Jahiruddin, M. (2016). Floodwater 301-303. https://doi.org/10.1038/537301a
ammonium, nitrogen use efficiency and rice yields Pardey, P. G., Andrade, R. S., Hurley, T. M., Rao, X., &
with fertilizer deep placement and alternate Liebenberg, F .G. (2016b). Returns to food and
wetting and drying under triple rice cropping agricultural R&D investments in Sub-Saharan Africa,
system. Nutrition Cycle of Agro-Ecosystem,104, 53-66. 1975–2014. Food Policy, 65, 1-8.
https://doi.org/10.1007/s10705-015-9758-6 https://doi.org/10.1016/j.foodpol.2016.09.009
Islam, S. M., Gaihre, Y. K., Biswas, J. C., Jahan, M. S., Singh, Parvan, A. (2011). Agricultural Technology Adoption: Issues
U., Adhikary, S. K., & Saleque, M. A. (2018). Different for Consideration When Scaling-Up.
nitrogen rates and methods of application for dry Ruzzante, S., Labarta, R., & Bilton, A. (2021). Adoption of
season rice cultivation with alternate wetting and agricultural technology in the developing world: a
drying irrigation: Fate of nitrogen and grain yield. meta-analysis of the empirical literature. World
Agricultural Water Management, 196,144-153. Development, 146, 105599.
https://doi.org/10.1016/j.agwat.2017.11.002 https://doi.org/10.1016/j.worlddev.2021.105599
Kinuthia, B. K. & Mabaya, E. (2017). The impact of Sadiq, M. S., Singh, I. P. & Ahmad, M. M. (2020). Rice yield
agricultural technology adoption on farmer welfare in differentials between IFAD participating and non-
Uganda and Tanzania. PEP Policy Brief No. 163. participating farmers in Nigeria’s Niger State.
Retrieved on 26/02/2022 from www.pep-net.org Economic Affairs, 65(4), 01-15.
(Partnership for Economic Policy) https://doi.org/10.46852/0424-2513.4.2020.12

54
Sadiq et al. (2022). Muş Alparslan University Journal of Agriculture and Nature, 2(2), 44-55

Sadiq, M. S., Singh, I. P., & Ahmad, M. M. (2021a). Navigating

fish food insecurity by simultaneous household and
marketed surplus-led productions in Kogi State of
Nigeria. Agricultural Socio-Economics Journal, 21(1),
41-50. https://doi.org/10.29103/aa.v8i2.4779
Sadiq, M. S., Singh, I. P., Ahmad, M. M., & Orifa, M. O.
(2021b). Impacts of farmers-herders conflict on
livelihoods of farming households in Nigeria’s
middle-belt region. Moroccan Journal of Agricultural
Science, 2(1), 14-24.
Wossen, T., Abdoulaye, T. Alene, A., Haile, G. M., Feleke, S.,
Olanrewaju, A., & Manyong, V. (2017). Impacts of
extension access and cooperative membership on
technology adoption and household welfare. Journal
of Rural Studies, 54, 223-233.
https://doi.org/10.1016/j.jrurstud.2017.06.022
Wossen, T., Berger, T., & Di Falco, S. (2015). Social capital, risk
preference and adoption of improved farm land
management practices in Ethiopia. Agricultural
Economics, 46(1), 81-97.
https://doi.org/10.1111/agec.12142
Zaidi, N. H., & Munir, A. (2014). Factors affecting the
adoption of agricultural technology in Bijnor District
of Western Uttar Pradesh. International Journal of Social
Science, 3(2), 205-216. https://doi.org/10.5958/2321-
5771.2014.00102.1

55