Digital Geography 3 (2022) 100027

Contents lists available at ScienceDirect

Digital Geography and Society

journal homepage: www.sciencedirect.com/journal/digital-geography-and-society

Application of the technology acceptance model to assess the use and

preferences of ICTs among field-level extension officers in Bangladesh
Khondokar H. Kabir a, *, Fuad Hassan b, Most. Zannatun Nahar Mukta a, Debashis Roy a,
Dietrich Darr c, Holli Leggette d, S.M. Asik Ullah a, e, f
a
Department of Agricultural Extension Education, Faculty of Agriculture, Bangladesh Agricultural University, Mymensingh 2202, Bangladesh
b
Department of Agricultural and Applied Statistics, Faculty of Agricultural Economics & Rural Sociology, Bangladesh Agricultural University, Mymensingh 2202,
Bangladesh
c
Rhine-Waal University of Applied Sciences, Faculty of Life Sciences, Germany
d
Department of Agricultural Leadership, Education and Communications, Texas A&M University, USA
e
Department of Environmental Design, Faculty of Design, Kyushu University, Japan
f
Department of International Studies, The University of Tokyo, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: Investment in ICTs has not always yielded the expected results due to inadequate usage of ICTs by extension
Extension and advisory service staffs. Our study aims to present a more general account of the use of ICTs and preferences of extension and
Information and communication technology advisory workers at the grassroots level. Following the technology acceptance model (TAM) and Dillman survey
(ICT)
method, a google form was used to collect data from 131 sub-district level extension officers covering 48 (out of
Technology acceptance model (TAM)
Dillman survey
64) districts of Bangladesh. Descriptive statistics and structural equation modeling was used to analyse the data
Structural equation modeling (SEM) and test the complex causal relationship of TAM components. Results revealed that the extension and advisory
Bangladesh workers obtained ICT tools from the extension organization but received insufficient organizational supports to
purchase internet packages, maintenance costs, and training to continue the use of it. Factor like perceived
usefulness of ICTs positively influenced the extension workers' intention to use ICT-based information systems. A
better understanding of the underlying reasoning behind the intention to use ICTs could aid future intervention
design and facilitate ICT adoption in extension services in Bangladesh and beyond.

1. Introduction communities to cover, which results in infrequent personal visits and

often lethargic service provision. The challenges became more apparent
Traditional extension and advisory services currently face significant after the onset of COVID-19. Mobility of advisors were temporarily
challenges in meeting farmers' diverse information needs and demands. suspended, and regular services were disrupted as tension among
The successful promotion of agricultural technology by providing farmers started rising. This issue is further compounded by the lack of
necessary information is essential to achieving the rural development human resources that limit the effectiveness and efficiency of advisory
goals (Mugwisi, Mostert, & Ocholla, 2015). A well-functioning extension services for rural farmers (Ragasa, Ulimwengu, Randriamamonjy, &
system is an important mechanism for disseminating information and Badibanga, 2015).
promoting the adoption of new farming technology among rural farmers One potential remedy for such situations is extension and advisory
(Suvedi, Ghimire, & Kaplowitz, 2017). The provision of quality exten­ workers' use of information and communication technologies (ICTs;
sion services significantly facilitates such adoption and brings changes Strong, Ganpat, Harder, Irby, & Lindner, 2014). ICTs are electronic tools
in agricultural productivity and farm income. In traditional extension for storing, processing, and disseminating information and communi­
systems, the extension and advisory staffs try to expose farmers to new cation (Balaji, Meera, & Dixit, 2007), which includes common
information and practices (Ghimire & Huang, 2015), but their effec­ communication tools like computers, mobile phones, and internet. Such
tiveness is limited by the large territory and numbers of farming tools now link rural people and extension staff with the necessary

* Corresponding author at: Department of Agricultural Extension Education, Bangladesh Agricultural University, Mymensingh, Bangladesh.
E-mail addresses: kabirag09@bau.edu.bd (K.H. Kabir), fuad.stat@bau.edu.bd (F. Hassan), zannatunnahar.agext@bau.edu.bd (Most.Z.N. Mukta), droyagext@bau.
edu.bd (D. Roy), dietrich.darr@hochschule-rhein-waal.de (D. Darr), hollileggette@tamu.edu (H. Leggette), asiksm@gmail.com (S.M.A. Ullah).

https://doi.org/10.1016/j.diggeo.2022.100027
Received 13 August 2021; Received in revised form 11 November 2021; Accepted 5 January 2022
Available online 7 January 2022
2666-3783/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

information and promote the diffusion of agricultural technology (Birke and advisory workers (Birke et al., 2019; Strong et al., 2014). The cur­
& Knierim, 2020; Birke, Lemma, & Knierim, 2019; Klerkx, Hall, & rent study aims to contribute to closing this gap.
Leeuwis, 2009). Moreover, digitization of farming, transition from me­ To study the perception of extension workers toward ICTs and their
chanical to data-based agricultural tools and practices, and multi-actor use of these technologies, we have selected Bangladesh as a case study.
cooperation can be useful in fostering agricultural innovation pro­ The government of Bangladesh has placed special emphasis on ‘digita­
cesses (Maria, Maria, & Andrea, 2021). Digital agricultural innovations lization’ as an instrument to achieve its Vision 2021 (Sakil, 2018). It
should be treated as a modern farm input to address communication stresses the importance of good governance and the free flow of infor­
challenges in agricultural advisory services (Steinke et al., 2020). Many mation to ensure better living conditions for the country's citizens. In
ICT-based development initiatives from around the world demonstrate line with the national vision, the Department of Agricultural Extension
the advantages and benefits of ICT in agricultural advisory service (DAE) has initiated efforts to use ICTs to provide real-time assistance to
provision, such as the Agriculture Knowledge Centers in Ethiopia (Birke the rural farmers and their advisors. DAE is Bangladesh's largest public
& Knierim, 2020); mobile phone-based animated videos in Burkina Faso agency under the Ministry of Agriculture and responsible for ensuring
(Maredia et al., 2018); the Agriculture portal in Bangladesh (DAE, government-led extension and advisory services to crop farmers and for
2020); or the IKisan portal in India (Aker, 2011). In the context of providing 24 million farm families with effective and secure extension
Europe, farmers and IT –experts made up the most common agricultural and advisory services based on their needs. Among the eight wings,
apps (Maria et al., 2021). According to Birke and Knierim (2020), such DAE's field service wing is responsible for providing farmers with
ICT-based initiatives are most effective, when they are properly sup­ advisory services at the grassroots level to improve their socio-economic
ported and administrated by extension and advisory staffs. However, in conditions. Yet, with an extension worker-to-farmer ratio of 3:7000,
practice their utilization by the extension and advisory staffs is often farmers often go long periods without seeing their assigned extension
poor (Strong et al., 2014). Hence, the effect of ICT on agriculture and and advisory workers. According to the DAE (http://www.dae.gov.bd
rural development, and the conditions necessary to benefit from these /site/page/7821c789-6cb4-46a9-9a57-e2af3d7880aa/-) and agricul­
technologies have attracted academic attention over the last decade ture portal (http://krishi.gov.bd/), approximately 15,000 extension
(Anastasios, Koutsouris, & Konstadinos, 2010; Sein, Thapa, Hatakka, & staffs provide 3.5 million Bangladeshi farmers with advisory services.
Sæbø, 2019). Studies found that the real use of ICTs in agricultural In the light of the facts, DAE recently introduced various ICTs to
extension and advisory services depends largely on agents' attitude to­ update its conventional advisory services (e.g., the agriculture portal
ward technology (Mugwisi et al., 2015; Mwansa, 2004). Growing evi­ (http://krishi.gov.bd/) and the mobile phone-based service ‘Krishak
dence from developing countries suggests that investment in ICTs do not Bandhu Phone Seba’) with a vow to extend e-agricultural services
automatically carry competitive benefits due to their inadequate usage throughout the country. DAE has also piloted various ICT devices (i.e.,
by the extension and advisory workers (see, Birke et al., 2019; Ong & Tablet and Internet) for their grassroots-level extension staff, which the
Lai, 2006; Strong et al., 2014). According to Steinke et al. (2020) National Agriculture Technology Program (NATP) funded (DAE, 2020).
pushing certain technologies rather than responding to particular These programs and devices are expected to enhance the extension
communication challenges of potential users is one important reason workers' professional engagement, job efficiency, and learning. Huge
why the extension system is not obtaining full benefit of digital media or investments have already been made to digitalize Bangladesh's advisory
ICTs. Extension officers' ability to effectively use various ICTs is services (i.e., web page development, capacity building training, expo­
dependent on continued funding, technical support, and training from sure visit, purchase and maintenance, and internet supports), yet in­
extension organizations (Sivakumar, Parasar, Das, & Anantharaman, vestment does not equal use.
2014). But the literature generally noted that extension employees in To the best of our knowledge, only one study has been conducted so
underdeveloped nations receive less support from their organizations far about the use and acceptance of ICTs by extension staff in
(see Smith & Kotrlik, 1990; Steinke et al., 2020). Existing literature in Bangladesh, but it particularly focused on social media (Kamruzzaman,
Bangladesh has also reported this problem. According to Kamruzzaman, Chowdhury, Paassen, & Ganpat, 2018). The purpose of this research was
Chowdhury, Odame, and Sarapura (2019), frontline extension workers to investigate the factors that could facilitate the use of Information
used social media and other ICT-based platforms to show their presence System (IS) by the extension and advisory staffs into their daily services.
to higher authorities. The reason for this pseudo-participation could be In the current paper, we address the research question: what are the
due to the system being pushed out of management without sufficient factors that affect extension and advisory workers' preferences and use
and ongoing organizational support. of information technology? Our research, therefore, aims to present a
Farmers' use of ICT-based apps, on the other hand, has frequently more general account of the use of ICTs and preferences of Bangladeshi
suffered as a result of their low educational and economic position. extension and advisory workers at the grassroots level. We advance the
Moreover, farmers may be unwilling to use ICT-based services due to current literature on ICT use among the agricultural extension and ed­
their technological incapability and habits, which may discourage ucation community in two ways. First, the study extends the Technology
extension workers from utilizing various ICTs (Steinke et al., 2020). Acceptance Model (TAM) by including some external variables in
Furthermore, ICT-based activities required positive impressions or re­ addition to the regular components of TAM to assess extension and
sponses from colleagues within the organization. If extension workers advisory workers' intentions to adopt ICTs. Second, our results have
who engage in ICT-based activities receive negative reactions from their direct policy implications as identifying the behavioral gaps in ICTs
coworkers, they may develop a negative attitude toward using it. In the adoption by field advisors can improve their integration in agricultural
context of Bangladesh, extension and advisory workers missed out on development strategies in developing countries like Bangladesh.
the opportunity to deliver a real-time response while engaging with
various stakeholders via social media because they prefer to push in­ 2. Theoretical discussion and hypothesis formulation
formation rather than engage in two-way interaction (Kamruzzaman
et al., 2019). They mostly prefer face-to-face connections with clients 2.1. Technology acceptance model (TAM)
because they perceive that this non-virtual engagement fosters strong
ties with farmers. However, although a growing body of literature has The technology acceptance model (TAM), developed in the 1980s
highlighted the importance of ICTs for effective agricultural extension (Davis, 1989), provides the conceptual underpinning of our research.
services for improved productivity, management and growth of the TAM was designed to explore the factors that shaped the intention of
value chain (Materia, Giarè, & Klerkx, 2015; Miller, Saroja, & Linder, workers from different sectors to use emerging information technologies
2013; Tata & McNamara, 2016), few studies have explored issues (IT), which should help organizations improve their workers' IT use.
related to the use of technology and the preferences held by extension TAM is derived from the theory of reasoned action (TRA), a theory to

2
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

describe individual's behavior and intention as they relate to a variety of authors classified each variable as ‘exogenous,’ ‘endogenous,’ and
activities (Fishbein & Ajzen, 1975). At the core of TRA is personal beliefs ‘moderating’ based on their role in the system. For example, organiza­
and subjective norms, which influence individual attitudes and expec­ tional support had been previously classified as an exogenous variable
tations that enhance their desire to carry out a behavior. that received no directional influence from other TAM variables even
Information system (IS) scholars, especially Davis (1989), used TRA though it may affect other TAM variables. Then, the ultimate intention
to explore the factors that determine the IT use behaviors of ultimate to use IT coined as endogenous variable linked to a one-directional
users (Holden & Karsh, 2010). Such research introduced several factors arrow and determined by exogenous variables. Finally, PU and PEOU
that formed the first version of TAM, as shown in Fig. 1 (Davis et al., were defined as moderating variables that received directional influence
1989). The original version of TAM had five components (i.e., perceived from another variable and also affected another variable of the system.
usefulness (PU), perceived ease of use (PEOU), attitude toward (AT), Strong et al. (2014) investigated technology use and preferences of
behavioral intention (BI), and actual use (U)). BI is mainly associated Caribbean extension officers. Their study used four constructs, including
with PU and PEOU, where PU is defined as the individual's belief that a technology usefulness, ease of use of technology, current use of tech­
particular IT would increase his or her job productivity, and PEOU is nology, and technology intentions. Strong et al. (2014) considered PU
defined as the individual's thinking about the extent of effort needed to and PEOU as two essential components of TAM that helped determine
use a specific IT (Davis, 1989). Over the years, TAM has developed into a extension staff's planned IT usage. In addition to the standard TAM
conceptual model widely used by scientists across contexts and enriched components, elements such as perceived responsiveness among farmers
by additional external variables (e.g., perceived responsiveness, orga­ and subordinates have been considered as important components in the
nizational support, task characteristics, job relevance, level of educa­ context of extension and advisory services.
tion, age, and prior similar experience (Amirtha & Sivakumar, 2018; In another study of TAM, Verma and Sinha (2018) identified the key
Legris, Ingham, & Collerette, 2003; Taherdoost, 2018). Legris et al. factors affecting rural farmers' adoption of mobile-based agricultural
(2003) conducted a comprehensive review to explore the external var­ extension services in India. Their study employed six constructs of TAM-
iables that added value to TAM in explaining IT use. They presented a perceived usefulness (PU), perceived ease of use (PEOU), social influ­
table with external variables considered in some studies and stated that ence (SI), attitude (At), perceived economic wellbeing (PEWB) and
they did not identify any clear pattern in selecting external variables for behavioral intention (BI). A contribution to the current TAM literature
TAM-related studies. They also noted in their findings that organiza­ from this research is that the perceived economic well-being precedes
tional support is more essential in predicting ICT use than user percep­ the perceived usefulness. However, they concluded all six constructs
tion of the technology. Amirtha and Sivakumar (2018) researched were reliable and valid.
whether the family life cycle stage affects the acceptability of e-com­ Finally, Almaiah, Jalil, and Man (2016) in their study considered
merce. Aside from the TAM model's core components, they used perceived responsiveness, in addition to standard TAM components, as a
generation-based grouping as an external variable and discovered that major external variable and found a positive impact on the respondents'
family life cycle stages significantly impact women's e-shopping perception of mobile learning application usefulness. Perceived
behavior in India. Taherdoost (2018) investigated how user acceptance responsiveness is defined as the degree to which people believe their
and recurrent usage of e-service technology is driven primarily by their partners are responsive in the professional relationship. In the context of
satisfaction with the system's quality and security. However, such the agricultural extension system, service providers are connected with
external variables have not been consistently documented in the liter­ their subordinates and with farmers. The positive response of farmers
ature as aiming to determine the intention of an adult to use IT. and subordinates to ICT-based initiatives affects the extension workers'
internal views about the PU of any ICT-based systems. Likewise, Pituch
and Lee (2006) reported that perceived responsiveness had an important
2.2. TAM in agricultural extension studies and positive effect on perceived usefulness of an e-learning program
among students in the US.
The TAM model has been widely used in agricultural studies (see, In this study, we consider perceived responsiveness as an external
Adnan, Nordin, & Bakar, 2017; Caffaro, Cremasco, Roccato, & Cavallo, variable in addition to the standard TAM variable. We also consider
2020; Flett et al., 2004; Musa, 2006; Sammah, Azril, Shaffril, Hassan, & organization variable as an external variable to provide a general
D'Silva, 2011; Zhang, Zhang, Mu, Zhang, & Fu, 2009), yet only few overview of extension and advisory workers' perceived beliefs about
aimed to identify factors influencing field extension and advisory how their organization supports the continuation of ICT-based extension
workers' use of IT. One of the notable exceptions was Sivakumar et al. service. However, a prior study focusing on software developers'
(2014) who explored the factors affecting the usage of field advisor perceived responsiveness revealed that it was an important external
computers in India. Their study incorporated individual-level factors (e. variable that strongly influenced both PU and PEOU of software users
g., age, education, annual income, job experience and gender) and (Gefen & Keil, 1998). This perceived responsiveness is equally important
organizational-level factors (e.g., organizational support for computer in the extension and advisory system, but to our knowledge, no study has
use beside the two main constructs PU and PEOU). In addition, the

Fig. 1. The Technology Acceptance Model, original version (Davis, Bagozzi, & Warshaw, 1989).

3
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

looked at this variable in the context of extension and advisory workers' 3. Method
usage of ICTs.
3.1. Study design and instruments of data collection
2.3. Formulation of hypotheses
The study used TAM which originated from a quantitative survey
Considering the above discussion, we considered exogenous vari­ study (Legris et al., 2003). Following the fact, this study also used a
ables (i.e., organizational support, and perceived responsiveness), quantitative survey research design. The quantitative survey method,
endogenous variables (i.e., current use of ICTs), and moderating vari­ rooted in empirical sociological research, appears to be well adapted for
ables (i.e., PU and PEOU) to develop the hypotheses (see Fig. 2). this study to explore the socio-psychological factors involved in tech­
In particular, the moderating variable i.e., the individual extension nology adoption. This study will improve understanding of the use and
and advisory worker's PU has been repeatedly demonstrated to influence preferences of information technology by extension and advisory staff.
the practical use of the IS (Igbaria, Parasuraman, & Baroudi, 1996; Based on the TAM model, we developed a questionnaire to investigate
Venkatesh & Davis, 2000). If the extension and advisory workers' belief extension and advisory officers' uses of and preferences for ICTs. The
that a particular IT would increase his or her job productivity, then it can questionnaire had two sections. Part A was a self-reporting tool con­
be assumed to increase the use of different IS. Therefore, it can be ex­ sisting of 19 items measuring the extension and advisory officers' de­
pected that: mographic information (4 items); current organizational support (10
items); and perceived responsiveness of farmers and subordinates in
Hypothesis A. PU is directly and positively related to IS use of the
ICTs use (5 items). Part B used the TAM instrument following the study
extension and advisory worker.
of Strong et al. (2014); Sivakumar et al. (2014); and Venkatesh and
Again, the individual extension and advisory worker's PEOU has Davis (2000) where 28 statements measured PU (9 items); PEOU (10
been repeatedly demonstrated to influence the PU (Arning & Ziefle, items); and actual use of ICTs (9 items). We measured farmers and
2007; Davis, 1989; Igbaria et al., 1996). If the extension and advisory subordinates' PR, PU, and PEOU through a four-point summated Likert
officers perceived that the use of a specific IT need minimum effort, then scale (1 = strongly disagree, 2 = disagree; 3 = agree, and 4 = strongly
it can be assumed to increase their belief on the usefulness of IS. disagree). Reporting on current use of ICTs was on a four-point rating
Therefore, it can be expected that: scale that considered ICT use in the week prior to the report (0 = Zero
times, 1–5 times; 6–10 times; and > 10 times). We measured all other
Hypothesis B. PEOU is directly and positively related to PU of the
items using nominal (1 = yes, 0 = no) and ratio scale.
extension and advisory worker.
Finally, practical usage of IS by farmers and subordinates has been 3.2. Population, sampling, and data collection
repeatedly shown to impact the PU and thus to affect the actual usage of
IS by extension and advisory service providers (Almaiah et al., 2016; DAE's field level extension and advisory workers were the population
Pituch & Lee, 2006). Farmers' and subordinates' enthusiastic behavior under study. In general, DAE's activities are decentralized at block,
about ICT-based programs may influence the internal beliefs of exten­ Upazila, district, and regional levels. DAE currently has 6818 officer
sion workers about the PU of any ICT-based systems. Therefore, it can be positions, which consists of 487 Upazila Agricultural Officers (UAO);
expected that: 487 additional agricultural officers (AAO); 974 agricultural extension
officers (AEO); 487 assistant agricultural extension officers (AAEO); and
Hypothesis C. PR of farmers and subordinates is directly and posi­
487 sub-assistant plant protection officers (SAAO) (DAE, 2020; also see
tively related to PU.
Fig. 3). The Upazila Agriculture Office, which is the administrative unit
In addition, any exogenous variable, i.e., organizational support for under the district level, is in charge of designing, conducting, moni­
the extension and advisory officers may influence the PU and PEOU toring and evaluating extension work. The SAAOs are under the direct
(Arning & Ziefle, 2007; Pijpers & Montfort, 2005). However, the in­ supervision of the Upazila agricultural officers. The government
clusion of such exogenous variables as organizational support into the launched a number of ICT-based initiatives to digitize the extension and
final model is highly conditional on validity tests. In general, organi­ advisory services at the Upazila level and instructed the extension and
zational support that aims to develop capacity among extension and advisory officers to become active in ICT-based initiatives (i.e., using
advisory workers can be assumed to increase their belief on IS and smartphones, internet, social media, video recording, and online
reduce the difficulties concerning the utilization of IS. reporting; uploading videos to the Agriculture Portal and the official
Facebook page; providing training to farmers and SAAOs; and moni­
toring activities in the field). Considering their roles and responsibilities,
our study explicitly focused on the Upazila extension personnel.
At the time of data collection, 981 extension personnel served at
Upazila level in 64 districts throughout Bangladesh with the remainder
of the positions being vacant. We purposively collected the official email
and phone number of all relevant extension and advisory officers from
the website of the Agricultural Information Service (http://www.ais.
gov.bd/). Using the collected contact information, we sent an email to
all extension and advisory officers requesting their participation in an
online survey hosted in a Google form (click here to see the Google
form). We notified participants that their data would be used for
research purposes only, and all personal identifiers would remain
confidential. We used Dillman et al.'s survey methods (Dillman, 2000) (i.
e., a systematic process of surveys and a series of reminders) to increase
the response rate.
We sent an email to the initial correspondence on August 1, 2018,
and we requested participants reply within two weeks. The initial cor­
respondence yielded a low response rate of 2.85% (only from 28 re­
Fig. 2. Conceptual Framework and Hypotheses. spondents) given that some of the provided email addresses were not

4
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

Fig. 3. Tiers of extension and advisory services including the number of workforces in Bangladesh.

accurate (number of delivery error was 54) and other explanations for 3.4. Data analysis
non-response (Nybakk, Crespell, Hansen, & Lunnan, 2009), such as lack
of time, indifference about the research, age, and education of the re­ We used statistical factorial analyses to test the study's hypotheses.
spondents. To increase the response rate, we randomly called re­ One of the forms of factorial analysis is exploratory factor analysis (EFA)
spondents who had not participated. We identified some additional which we employed to identify the latent constructs in the data and
reasons for disinterest, including the respondents' lack of access to the ascertain the factorial validity of these constructs (Hair, Anderson,
official email address, lack of online survey experience, and unfamil­ Tatham, & Black, 1998; Lee, Hsieh, & Chen, 2013). Because the latent
iarity with the research group. By the end of November 2018, 131 constructs in social science are often interrelated, we used the maximum
extension and advisory workers from 105 Upazilas in 48 districts had likelihood (ML) factor extraction method with promax rotation to
responded, covering 13% of the total population of Upazila-level permit inter-factor correlation. We selected the factors with an eigen­
extension workers, 22% of the total Upazilas, and 75% of the total dis­ value exceeding one for rotation (Field, 2000). Because the sample size
tricts in Bangladesh (Fig. 4). We measured the internal consistency and was 131, a factor loading of 0.131 was used as a lower cut-off value for
reliability by using Cronbach alpha (α). Additionally, we calculated the selection of variables for each factor (Field, 2000). The latent factors
convergent validity measures (i.e., item-to-item and item-to-total cor­ identified through EFA were subjected to confirmatory factor analysis
relations; Doll & Torkzadeh, 1988). (CFA) to verify the factor structure and estimate both reliability and
validity. Finally, we later tested the hypothesized relationships in the
3.3. Description of the sample research model through structural modeling. We estimated reliability
estimation, exploratory factor analysis (EFA), confirmatory factor
A total of 131 extension and advisory workers completed a Google analysis (CFA), and structural modeling through STATA (Version 15).
form to provide data. Table 1 represents the characteristics of the re­
spondents. The respondents' age ranged from 24 to 50 years with a mean 4. Results and discussion
of 32.3 years. The majority (75.57%) of the respondents were in the age
group of 24 to 35 years. More than half (55.73%) of the respondents 4.1. Use of information and communication technology (ICT)
were agricultural extension officers (AEO) with 44.27% being Upazila
agriculture officers (UAO). An overwhelming 86.26% of the respondents The results shown in Table 2 indicate that respondents used ICTs
had a post-graduate degree (i.e., masters). More than half of the re­ more frequently for personal contact (M = 3.59; SD = 0.760), personal
spondents (53.44%) had professional experience of one to five years. tasks (M = 3.47; SD = 0.764), communicating with supervisors and
According to the respondents' profile, DAE's entry level officers (i.e., colleagues (M = 3.30; SD = 0.835), and personal knowledge enhancing
AEOs) had a relatively higher response rate to an ICT-based data gath­ (M = 3.23; SD = 0.873). The mean score indicates that respondents used
ering instrument than the higher-ranked officers (i.e., UAOs). the ICTs 6–10 times per week. Conversely, respondents used ICTs less
frequently to prepare extension teaching materials (M = 2.52). Mean
values (M = 2.628) clearly show that respondents used ICTs less

5
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

Fig. 4. Figure showing 105 responded upazila, covering 13% of the total extension workers in Bangladesh.

frequently to store extension teaching materials and to provide advisory omnipresent in almost every professional and personal context. A recent
services to farmers. On average, respondents used ICTs between 1 and 5 study by Islam, Haque, Afrad, Abdullah, and Hoque (2017) on
times per week for the above noted purposes. Thus, respondents used Bangladesh extension professionals found that DAE's field-level officers
ICTs more for personal purposes than they did for enhancing their used cell phones more frequently for personal purposes (100%) than for
professional work in extension. official purposes (93.64%). Also, Mukta, Karim, and Kashem (2010)
This result, however, is not surprising because ICTs are increasingly found farmers in Bangladesh increasingly used cell phones. Likewise, a

6
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

Table 1 Table 3
Respondents' Profile (N = 131). Indicators of reliability and convergent validity.
Characteristics n % Cumulative Scale Item Cronbach's Convergent validity
α
Age Item-test Item-rest
24–29 years 44 33.59 33.59 correlation range correlation
30–35 years 55 41.98 75.57 range
36–40 years 19 14.50 90.07
Information 9 0.84 0.46–0.83 0.31–0.76
41–45 years 9 6.87 96.94
system use
46–50 years 4 3.05 100.00
PR 3 0.74 0.68–0.70 0.26–0.34
Working Position
PU 8 0.76 0.43–0.62 0.23–0.38
UAO 58 44.27 44.27
PEU1 5 0.71 0.51–0.64 0.22–0.44
AEO 73 55.73 100.00
PEU2 4 0.51 0.50–0.66 0.12–0.21
Education
OS 3 0.39 0.42–0.65 0.11–0.28
Bachelor 18 13.74 13.74
Masters 113 86.26 100.00
Job Experience
1–5 years 70 53.44 53.44 required level of 0.70 for any further analysis (Nunnally & Bernstein,
6–10 years 45 34.35 87.79 1994). There are no hard and fast rules for justifying the validity and
11–15 years 9 6.87 94.66 consistency, but there are rules of thumb for minimally required
16–20 years 7 5.34 100.00 magnitude to accept the outcomes. According to Robinson, Shaver, and
Note: UAO = Upazila Agriculture Officer; AEO = Agricultural Extension Officer. Wrightsman (1991), the convergent validity is achieved if the sub-scale
item-rest correlations exceed 0.50 and inter-item correlations exceed
0.30. However, a recent study of Zijlmans, Tijmstra, Van der Ark, and
Table 2 Sijtsma (2018) showed that minimally required values of item-rest
Average frequency of ICT use by extension officers (N = 131). correlations are 0.20, 0.30, or 0.40 for maximum-performance tests.
Use of ICTs M SD The results displayed in Table 3 show that most of the sub-scales had
Personal contact 3.586 0.760 adequate convergent validity. But the Cronbach's alpha, item-test and
Enhance your personal knowledge 3.231 0.873 item-rest of PEU2 and OS are below the minimum required average.
Personal tasks 3.471 0.764 That's why OS and PEU2 are eliminated for further causal relationship
Contact farmers and give your advice 2.619 0.985 analysis but considered for descriptive statistical analysis. Following the
Communicating with supervisors and colleagues 3.305 0.835
Acquire the knowledge to enhance your training 2.867 0.999
initial criteria, we collected further data for an exploratory factor
Search for information when preparing your programs 2.892 0.901 analysis (EFA) to identify the latent factor structure in the data. In fact,
Store teaching materials 2.628 1.025 EFA is used to get the unique and uncorrelated items from correlated
Prepare training materials 2.520 1.025 items in the data set.

recent study showed that American citizens on average spend 231 min 4.3. Insights from empirical data on organizational supports
on mobile internet and 50 min on computer internet every day, totaling
more than 4.5 h of ICTs use per day for personal contexts (Molla., 2020). Organizational support in the form of financial, technical, or other
An increased use of ICTs can enhance the capacity and convenience of motivational incentives are influential factors that enable workers to use
using such tools for more technical purposes, including professional use. ICTs in the provision of extension service (Sivakumar et al., 2014;
Strong et al. (2014) stated that Caribbean extension workers used ICTs Venkatesh & Davis, 2000; Yi & Hwang, 2003). Gow, Chowdhury,
for personal purposes, which helped them develop professional capacity. Ramjattan, Ganpat, and Gow (2020) observed that formal in-service
A recent study in the context of Bangladesh found that extension training in technology stewardship can greatly contribute to extension
agents perceive social media as a means for improving their professional workers' willingness to implement their learning in a community of
performances (Kamruzzaman et al., 2018). Likewise, findings of Kam­ practice as a post-course activity. Therefore, although the items in
ruzzaman et al. (2019) indicate that the extension agents use social organizational support did not include in the structural equation model,
media to learn and improve professional tasks and performances and it seems important to address the variable from the empirical evidence
reflect on those activities as a problem-solving process to gain short term and overall research insights. The empirical data regarding organiza­
tangible outcomes. However, the extension personnel rarely use social tional support is presented in Table 4. The results showed insufficient
media for critical discussion and interaction with varied stakeholders to organizational support for extension workers who use ICTs at the field
initiate changes in traditional activities of the extension organization to level. Most extension professionals (83.97%) did not receive monthly
obtain long term benefits (Kamruzzaman et al., 2019; Kantamara & budgets to cover their costs of internet access, and 64.12% indicated that
Vathanophas, 2014). The study by Steinke et al. (2020) supports the fact they had no management budget for ICT devices. In addition, approxi­
that the extension and advisory workers initially adopt ICT-based ini­ mately 57.25% of the employees have received ICT tools (e.g., tablet,
tiatives due to pressure from higher authority but discontinue at the end laptop) from DAE, but most respondents (67.94%) have not been trained
of the project because they are not fully aware of the full potential of
digital initiatives to improve the performance of extension system. In the Table 4
context of Bangladesh, the level of ICT usage by extension officers may Organizational support (N = 131).
be influenced by many factors as theorized in the TAM. We explore these
Types of support Response F %
variables and their effects on ICT use in the following sub-sections.
Receive a monthly budget to buy internet package for No 110 83.97
official use Yes 21 16.03
Receive management budget for office to set-up and No 84 64.12
4.2. Reliability and validity of the factor analysis maintenance of ICTs tools Yes 47 35.88
No 89 67.94
Receive ICTs related training in the last six months
The Cronbach's alpha shows how consistently individuals respond to Yes 42 32.06
the items within a scale or sub-scale (Nunnally & Bernstein, 1994) Receive any ICT tools (e.g., tab, laptop, modem, No 56 42.75
projector) from DAE Yes 75 57.25
measured from the pilot data. Usage of information system showed was
adequate (see Table 3) as the reliability alpha value exceeded the Note: DAE = Department of Agricultural Extension.

7
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

on ICTs in the last six months. Table 5

The DAE offered ICT resources to help over half of the respondents, Results of exploratory factor analysis.
but DAE missed out on the requisite training and opportunities. At Variable ICT PEU1 PU PR
Upazila level, extension workers (Fig. 3) engaged mainly with small
1 B38 0.8824
groups of farmers pursuing specific projects. DAE expects its officials to 2 B39 0.8590
use ICTs to provide a large number of farmers with need-based advisory 3 B37 0.7040
services. Officers are rarely inspired to keep spending money on mobile 4 B32 0.6762
bills and internet packages to exchange knowledge with farmers. In 5 B34 0.5862
6 B33 0.5619
addition, the vast majority of DAE officials at middle to old age ranges 7 B35 0.4392
are used to traditional extension and advisory services (SAAOs, Fig. 3). 8 B36 0.4297
Without training, financial rewards, and continued support from the 9 B31 0.3423
government, extension workers may not consider using ICTs to be 10 B24 0.7506
11 B28 0.5662
convenient, interesting or useful. A study by Tata and McNamara (2018)
12 B26 0.4815
stated that qualified extension workers were more likely than the un­ 13 B22 0.3953
trained staff to use the computer system in extension work. 14 B210 0.3626
Our results have, however, suggested a void in DAE's strategy and 19 B12 0.7045
action to provide sufficient organizational support to optimize ICT in­ 20 B14 0.6104
21 B17 0.5844
formation system usage during extension work. This finding is corrob­ 22 B13 0.4078
orated by results illustrating the importance of organizational support 23 B15 0.3919
for ICT adoption from other economic sectors. For example, Hoque, Saif, 24 B18 0.3829
AlBar, and Bao (2016) found that government, management, and 25 B19 0.3778
26 B11 0.3529
financial support were important determinants of the adoption of ICTs in
27 B16 0.2076
various SMEs in Bangladesh. Management support and infrastructure 32 A41 0.3922
have also been identified as important factors for ICTs adoption in India 33A44 0.4816
(Todd & Javalgi, 2007) and Europe (Bayo-Moriones & Lera-López, 34 A45 0.3909
2007). 35 A43 0.2974
36 A42 0.2896
Variance explained (%)
4.4. Identifying latent variables through exploratory factor analysis 32.91 15.61 11.17 8.30
Eigen Value
After testing initial reliability and validity, we identified the latent 4.63 2.20 1.29 1.10
factor structure in the data using an EFA. We used the method of
maximum likelihood to identify factors and pattern matrix. The deter­
confirmatory factor analysis is iteratively fit after several analyses based
minant of the correlation matrix is significant. The estimate of the
on the estimates and goodness-of-fit indicators. The final confirmatory
Kaiser-Meyer-Olkin (KMO) value of 0.725 was above the minimum level
factor analysis model along with estimated unstandardized regression
of 0.5 (Kaiser, 1974). All the initial criteria (i.e., a significant determi­
coefficients is displayed in Fig. 5.
nant of the correlation matrix [0.0001], KMO value [0.725], and a
The results show that the specified model has Likelihood ratio Chi-
significant Bartlett's Chi-square [χ2 = 1288.615, p < 0.001]) indicate
square 350.606, which is significant at 5%. The first chi-square re­
that the sample size chosen for this investigation was adequate (Field,
ports the model versus saturated test and the second one is baseline
2000) and sufficient inter-correlation exists to run the factorial analysis.
versus saturated test. The saturated model fits the covariance perfectly.
We initially identified six factors but only retained four factors for
However, for an adequate fit, the Chi-square test statistic should be
further analysis, which surpassed the eigenvalue of one for rotation.
insignificant. The root mean squared error of approximation (RMSEA)
Together, these factors explained 87.46% of the total variance in the
and p close for a test also indicated a close fit (Table 6). The lower bound
data, which is above the acceptable level of 60% (Hair et al., 1998). The
(0.032) and upper bound (0.063) of the 90% confidence interval indi­
pattern matrix is displayed in Table 5, showing that all the questionnaire
cate the model is a close fit. According to Browne and Cudeck (1989), if
items (numbered) related to the observed item loaded exactly (values
the lower bound of the 90% confidence interval is below 0.05 and upper
exceeding 0.30 on a single factor) onto their hypothesized factors. Most
bound does not exceed 0.10, then the model fit is close. The Akaike
commonly recommended minimum item-factor loadings are 0.3 to 0.4
information criterion (AIC) and Bayesian (or Schwarz) information cri­
(Gorsuch, 1983; Nunnally, 1978; Tabachnick & Fidell, 2007). We
terion (BIC) statistics compare the fit of different models. Smaller values
labeled items B31 to B39 on factor 1 as ICT, and nine defined items in the
indicate a better fit (Kline, 2016). Some of the item loadings are dropped
first factor explained 29.10% of the variance in ICT. For example, B22,
and added, and the combination of the same categorized latent variable
B24, B26, B28, B21 were part of the perceived ease of use-1 (PEU1)
as PEU1 and PEU2 are compared with these statistics. Comparative fit
latent factor. PEU1 indicates easiness of technology use and PEU2 in­
index (CFI) and Tucker–Lewis index (TLI) are two indices such that a
dicates the opposite. The items loaded in the second factor named PEU1
value close to 1 indicates a good fit (Bentler, 1990; Hair et al., 1998).
explained 15.61% of the variance. B11, B12, B13, B14, B15, B17, B18,
The standardized root mean squared residual (SRMR) (SRMR = 0.082)
and B19 represented the latent factor perceived usefulness. The loaded
reports a good fit of the model. A good fit is a small value, considered by
items in the responsiveness of use explained 11.17% in each latent
some to be limited to 0.08 (Hancock & Mueller, 2013). The coefficient of
factor. The loadings of the items ranged from 0.20 to 0.88. However, all
determination (CD = 0.791) also indicates a good fit. For a perfect fit, CD
the items were not significantly loaded for relevant latent construct.
should correspond to a value of 1.0.

4.5. Verifying factor structure through confirmatory factor analysis

4.6. Structural equation model (SEM) and hypotheses test
The latent factor structure identified through exploratory factor
analysis allowed us to perform confirmatory factor analysis to enhance Table 7 represents the standardized coefficient of the parameters,
its convergent validity. We omitted several items in PEU and PR for the standard error and the confidence interval of the coefficients. Based on
validity construct and the final iterative fitting process as they did not the result of the structural equation model, the estimated path co­
substantially load into the specified factor and were less than 0.31. The efficients (Fig. 6) demonstrate that the perceived responsiveness (path

8
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

Fig. 5. Confirmatory factorial analysis.

coefficient = 0.128) was directly and positively related with perceived Our results also suggest that the PEU had an indirect influence on the
usefulness and thus indirectly related with the information system use. actual use of ICTs (Fig. 6). Other research in the context of Bangladesh
Thus, the study results support hypothesis C: perceived responsiveness of revealed a similar trend between the PEU and actual use of ICTs. For
farmers and subordinates is directly related with perceived usefulness. instance, Kamruzzaman et al. (2018) found that the perceived ease of
According to the estimated path coefficient value in Fig. 6, the use positively influenced the use of social media by the Bangladesh
perceived ease of use (path coefficient = 0.141) had a direct and positive extension professionals. Furthermore, Hamid, Razak, Bakar, and
relationship with the perceived usefulness, which supported the hy­ Abdullah (2016) reported a correlation between PEU and use of e-gov­
pothesis B: perceived ease of use is directly and positively related with ernment services.
perceived usefulness. The finding is consistent with past studies of Siva­ As hypothesized, the PU of ICTs, with a path coefficient of 0.515, had
kumar et al. (2014), Venkatesh and Davis (2000) and Yi and Hwang strong direct and positive effects on the information system use. Thus,
(2003). For example, Gurtner, Reinhardt, and Soyez (2014) found that our results support hypothesis A: perceived usefulness is directly and
the PEU influenced PU for adoption of mobile-based applications in positively related with information system use. The result suggests that, if
Germany. A similar influence was found in mobile banking apps field-level advisors view ICTs as helpful in enhancing their job perfor­
(Munoz-Leiva, Climent-Climent, & Leibana-Cabanillas, 2017 as cited in mance, their intention to use these technologies increases. Our research
Verma & Sinha, 2018, p-4). results support the work of Hamid et al. (2016), Sivakumar et al. (2014),

9
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

Table 6 Bangladesh extension professionals' use of ICTs (Table 7).

Goodness of fit of the model.
Fit statistic Value Description ▪ Extension workers' perceived usefulness of ICTs had a direct,
positive, and statistically significant influence on the use of the
Likelihood ratio
χ2_ms (272) 350.606 model vs. saturated information system. Hence, we can confirm Hypothesis A.
p > chi2 0.001 ▪ Extension workers' perceived ease of use of ICTs was directly and
χ2_bs (300) 857.414 baseline vs. saturated positively, but not significantly, related with their perceived
p > chi2 0 usefulness of ICTs. Hence, we cannot confirm Hypothesis B.
Population error
RMSEA 0.049 Root mean square error of approximation
▪ Farmers and subordinates' perceived responsiveness toward ICT
90% CI, lower bound 0.032 use had influenced extension workers' perceived usefulness of
upper bound 0.063 ICTs positively but not at statistically significant levels. Hence,
P close 0.539 Probability RMSEA ≤ 0.05 we confirm Hypothesis C. The results support Sivakumar et al.
Information criteria
(2014) explanation of variables (i.e., direction of relationship
AIC 7362.309 Akaike's information criterion
BIC 7580.381 Bayesian information criterion with other variables of TAM).
Baseline comparison
CFI 0.859 Comparative fit index The goal of this study was to use extended TAM as a theoretical
TLI 0.844 Tucker-Lewis index framework to investigate the factors that influenced extension and
Size of residuals
SRMR 0.082 Standardized root mean squared residual
advisory workers' willingness to use emerging ICTs. We included
CD 0.791 Coefficient of determination perceived responsiveness and perceived ease of use in the TAM model.
These two variables appear to be predictors of the perceived effective­
ness of ICTs among extension and advisory professionals. Finally, this
Table 7
study demonstrated that both variables have no relation to the intension
The results of the structural equation model. to use ICTs. Rather, this study finds that the perceived usefulness of ICTs
positively influenced the extension workers' intension to use ICT-based
Coefficient Std. Z P>z [95% Confidence
Err. Interval]
information systems.

Lower Upper
bound bound
4.7. Limitations and future research scope
Usefulness
Responsiveness 0.128 0.174 0.736 0.461 − 0.213 0.470 Our research has three drawbacks that should be considered in future
Ease of use 0.141 0.144 0.979 0.327 − 0.141 0.423
ICT
research to provide further insight into the subject area.
Usefulness 0.515 0.0954 5.398 0.000 0.328 0.702 First, the items and measurement scales for the organizational support
variable did not fit the explanatory factor analysis and, as a result, were
not considered for the structural equation model. The analysis of
Venkatesh and Davis (2000) and Ong and Lai (2006). Sivakumar et al. empirical data, however, suggested insufficient organizational support
(2014) reported that, when extension workers recognize the usefulness to improve the DAE officials' use of ICT-based information systems. In
of computers in extension work, they are likely to put more effort in this regard, an in-depth qualitative investigation, supported by quanti­
computer system use. Additionally, Davis (1989) also explained the aims tative research, should be conducted to provide a complete picture
to use computer systems by the people when they see a chance of showing how the organizational support system plays a role in exten­
enhancing their professional capacity using them. Furthermore, Azam sions' ICT use.
and Quaddus (2013) in Bangladesh found that certain cultural variables, Second, although we invited the complete population of Upazila-
users' PEU and PU have significant effects on users' intention to use ICTs level extension officers to participate in our study, we had a small
in Small and Medium Enterprises (SMEs). Similarly, other studies sample size. Extension officers who responded were interested or
revealed that the PU predicted the intention of mobile internet use (Kim, motivated in the topic, which could have led to sample selection bias.
Chan, & Gupta, 2007), smartphone use by doctors (Park & Chen, 2007); DAE officers' intention to use ICTs at field level may, therefore, be
and mobile learning adoption (Callum, Jeffrey, & Kinshuk., 2014 as overestimated in our study.
quoted in Verma & Sinha, 2018, p. 4). Third, only a few external and moderating variables from the TAM
Based on the Structural Equation Model (SEM), the hypotheses model are included in our study. Bangladesh has distinct socio-political
confirmed the following directional influences among the various and cultural identities as a ‘developing’ South Asian nation (Gupta,
exogenous, moderating, and endogenous variables relating to 2012). Thus, cultural variables, which are not explored in the current

Fig. 6. Estimated path coefficient of the SEM.

10
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

study, can also influence people's perception, desire, and interaction interests or personal relationships that could have appeared to influence
through ICT (Azam & Quaddus, 2013). Therefore, the behavior of the work reported in this paper.
extension workers in rural Bangladesh should also involve culturally
specific variables in future studies. References

5. Conclusions and recommendations Adnan, N., Nordin, S. M., & Bakar, Z. B. A. (2017). Understanding and facilitating
sustainable agricultural practice: A comprehensive analysis of adoption behaviour
among Malaysian paddy farmers. Land Use Policy, 68, 372–382. https://doi.org/
With the continuous evolution of communication technology and 10.1016/j.landusepol.2017.07.046
related shifts in human behavior, the use of ICTs in extension and Aker, J. C. (2011). Dial “A” for agriculture: A review of information and communication
technologies for agricultural extension in developing countries. Agricultural
advisory services is becoming more relevant. DAE has launched several Economics, 42(6), 631–647. https://doi.org/10.1111/j.1574-0862.2011.00545.x
ICT-based initiatives in Bangladesh enabling their traditional services to Almaiah, M. A., Jalil, M. A., & Man, M. (2016). Extending the TAM to examine the effects
be turned into a digitized form. In many cases, mobile phones, social of quality features on mobile learning acceptance. Journal of Computers in Education,
3(4), 453–485. https://doi.org/10.1007/s40692-016-0074-1
media, and other internet-based interactions replace traditional one- Amirtha, R., & Sivakumar, V. J. (2018). Does family life cycle stage influence e-shopping
way communication channels (e.g., radio, television). Such ICTs pro­ acceptance by Indian women? An examination using the technology acceptance
vide space for bidirectional engagement of rural populations. In agri­ model. Behaviour & Information Technology, 37(3), 267–294. https://doi.org/
10.1080/0144929X.2018.1434560
cultural extension services, the smooth operation of various ICTs mainly
Anastasios, M., Koutsouris, A., & Konstadinos, M. (2010). Information and
depends on how the field-level extension workers perceive and use these communication technologies as agricultural extension tools: A survey among farmers
technologies. Our study investigated the field-level extension and in West Macedonia, Greece. The Journal of Agricultural Education and Extension, 16
(3), 249–263. https://doi.org/10.1080/1389224X.2010.489767
advisory workers' use of ICTs through the lens of TAM.
Arning, K., & Ziefle, M. (2007). Understanding age differences in PDA acceptance and
The results of our study support the hypothesis that the extent of performance. Computers in Human Behavior, 23(6), 2904–2927.
professional use of ICTs has been primarily affected by the employees' Azam, M., & Quaddus, M. (2013). Examining the influence of national culture on
perception about the usefulness of technology to strengthen the exten­ adoption and use of information and communication technology: A study from
Bangladesh’s SME perspective. The International Technology Management Review, 3
sion advisory services. Results also showed that farmers and sub­ (2), 116–126.
ordinates' PR and officers' PEU of ICT had some positive, but indirect Balaji, V., Meera, S. N., & Dixit, S. (2007). ICT-enabled knowledge sharing in support of
effects on their ICTs usage. However, empirical results also indicated extension: Addressing the agrarian challenges of the developing world threatened by
climate change, with a case study from India. SAT eJournal, 4(1), 1–18.
that extension staff's technical use of ICTs was lower than their personal Bayo-Moriones, A., & Lera-López, F. (2007). A firm-level analysis of determinants of ICT
use of ICTs. In addition, a low survey response rate suggested that adoption in Spain. Technovation, 27(6–7), 352–366. https://doi.org/10.1016/j.
extension staff were not interested in using ICTs for professional and technovation.2007.01.003
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin,
networking purposes. Nevertheless, high levels of personal ICTs usage 107(2), 238–246. https://doi.org/10.1037/0033-2909.107.2.238
can also suggest a steady increase in ICT usage efficiency for extension Birke, F. M., & Knierim, A. (2020). ICT for agriculture extension: Actor network theory
and advisory services provision in the medium term. for understanding the establishment of agricultural knowledge centers in south
Wollo, Ethiopia. Information Technology for Development, 26(3), 591–606. https://
The study indicated a distinct gap between DAE's plan and actions
doi.org/10.1080/02681102.2020.1727826
regarding the digitization of the extension system. Our study found that Birke, F. M., Lemma, M., & Knierim, A. (2019). Perceptions towards information
current DAE support is insufficient to mobilize extension staff's use ICTs communication technologies and their use in agricultural extension: Case study from
and enhance the delivery of extension services in rural areas. Although south Wollo, Ethiopia. The Journal of Agricultural Education and Extension, 25(1),
47–62. https://doi.org/10.1080/1389224X.2018.1524773
over half of the respondents obtained ICT tools from DAE, they were not Browne, M. W., & Cudeck, R. (1989). Single sample cross-validation indices for
consistently supported by DAE in using these devices. Field-level covariance structures. Multivariate Behavioral Research, 24(4), 445–455. https://doi.
extension practitioners are often criticized by their clients for failing org/10.1207/s15327906mbr2404_4
Caffaro, F., Cremasco, M. M., Roccato, M., & Cavallo, E. (2020). Drivers of farmers’
to offer face-to-face advisory services to overpopulated farm commu­ intention to adopt technological innovations in Italy: The role of information
nities, but on the contrary, DAE officials at the national level continually sources, perceived usefulness, and perceived ease of use. Journal of Rural Studies, 76,
urge their field staff to increase their ICT use despite being provided with 264–271. https://doi.org/10.1016/j.jrurstud.2020.04.028
Callum, K. M., Jeffrey, L., & Kinshuk.. (2014). Comparing the role of ICT literacy and
little or insufficient incentive and training. Based on our results of anxiety in the adoption of mobile learning. Computers in Human Behavior, 39, 8–19.
extension workers' socio-demographic characteristics, we conclude that https://doi.org/10.1016/j.chb.2014.05.024
mostly younger extension officers are trying to navigate this paradoxical Davis, F. (1989). Perceived usefulness, perceived ease of use, and user acceptance of
information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/
situation by using ICTs in their profession. In the light of the findings of 249008
this study, DAE should expand and continue to provide advanced ICT Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1989). User acceptance of computer
training to its field staffs so they can use ICT for professional purposes technology: A comparison of two theoretical models. Management Science, 35(8),
982–1003.
effectively. The trainings should aim to inspire field staffs to recognize
Department of Agricultural Extension (DAE). (2020). Agricultural Extension Manual.
and view ICTs as easy and useful resources for developing extension and Ministry of Agriculture, Government of the People's Republic of Bangladesh.
advisory services across digital platforms. DAE should reinforce the Retrieved from: http://www.dae.gov.bd/site/page/7821c789-6cb4-46a9-9a57-e2af
current support structures to ensure the continuous use of ICTs at the 3d7880aa/.
Dillman, D. A. (2000). Internet and mail surveys: The tailored design method. New York:
organizational level. The design of an intervention strategy for projects Wiley.
controlled by the DAE should include adequate incentives, logistical Doll, W. J., & Torkzadeh, G. (1988). The measurement of end-user computing
support, and ICT maintenance. Finally, DAE should encourage field of­ satisfaction. MIS Quarterly, 12(2), 259–274.
Field, A. (2000). Discovering statistics using SPSS for windows: Advanced techniques for the
ficials to expand their network and to collaborate with other develop­ beginner. London: Sage.
ment partners/stakeholders through the ICT use. In this regard, DAE Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to
should create opportunities for those who make substantial contribu­ theory and research. Reading: Addison-Wesley.
Flett, R., Alpass, F., Humphries, S., Massey, C., Morriss, S., & Long, N. (2004). The
tions to advisory services and achieve success through the creative use of technology acceptance model and use of technology in New Zealand dairy farming.
ICTs (in the form of award or special bonus). This will encourage Agricultural Systems, 80(2), 199–211. https://doi.org/10.1016/j.agsy.2003.08.002
extension staffs to use ICTs in extension teaching, training material Gefen, D., & Keil, M. (1998). The impact of developer responsiveness on perceptions of
usefulness and ease of use: An extension of the technology acceptance model. Data
preparation, and sharing those with target clients, colleagues and rele­ Base for Advances in Information Systems, 29(2), 35–49. https://doi.org/10.1145/
vant development partners. 298752.298757
Ghimire, R., & Huang, W.-C. (2015). Household wealth and adoption of improved maize
varieties in Nepal: A double-hurdle approach. Food Security, 7(6), 1321–1335.
Declaration of Competing Interest https://doi.org/10.1007/s12571-015-0518-x
Gorsuch, R. (1983). Factor analysis (3rd ed.). Hillsdale, NJ: Lawrence Erlbaum.
The authors declare that they have no known competing financial

11
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

Gow, G., Chowdhury, A., Ramjattan, J., Ganpat, W., & Gow, G. (2020). Fostering Musa, P. F. (2006). Making a case for modifying the technology acceptance model to
effective use of ICT in agricultural extension: Participant responses to an inaugural account for limited accessibility in developing countries. Information Technology for
technology stewardship training program in Trinidad. The Journal of Agricultural Development, 12(3), 213–224. https://doi.org/10.1002/itdj.20043
Education and Extension, 26(4), 335–350. https://doi.org/10.1080/ Mwansa, D. M. (2004). Demystifying learning and knowledge: Extending the scope of
1389224X.2020.1718720 literacy. Adult Education and Development, 61, 87–98.
Gupta, A. (2012). Pesticide use in south and South-East Asia: Environmental public Nunnally, J. (1978). Psychometric theory (2nd ed.). New York, NY: McGraw-Hill.
health and legal concerns. American Journal of Environmental Sciences, 8(2), 152–157. Nunnally, J. C., & Bernstein, I. H. (1994). The assessment of reliability. Psychometric
Gurtner, S., Reinhardt, R., & Soyez, K. (2014). Designing mobile business applications for Theory, 3, 248–292.
different age groups. Technological Forecasting and Social Change, 88, 177–188. Nybakk, E., Crespell, P., Hansen, E., & Lunnan, A. (2009). Antecedents to forest owner
https://doi.org/10.1016/j.techfore.2014.06.020 innovativeness: An investigation of the non-timber forest products and services
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data sector. Forest Ecology and Management, 257, 608–618. https://doi.org/10.1016/j.
analysis (5th ed.). Upper Saddle River, NJ: Pearson Education. foreco.2008.09.040
Hamid, A. A., Razak, F. Z. A., Bakar, A. A., & Abdullah, W. S. W. (2016). The effects of Ong, C., & Lai, J. (2006). Gender differences in perceptions and relationships among
perceived usefulness and perceived ease of use on continuance intention to use e- dominants of e-learning acceptance. Computers in Human Behavior, 22(5), 816–829.
government. Procedia Economics and Finance, 35, 644–649. https://doi.org/10.1016/ Park, Y., & Chen, J. V. (2007). Acceptance and adoption of the innovative use of
S2212-5671(16)00079-4 smartphone. Industrial Management and Data Systems, 107(9), 1349–1365.
Hancock, G. R., & Mueller, R. O. (2013). Structural equation modeling: A second course. Pijpers, G. G. M., & Montfort, K. V. (2005). An investigation of factors that influence
Charlotte, NC: Information Age Publishing. senior executives to accept innovations in information technology. International
Holden, R. J., & Karsh, B. T. (2010). The technology acceptance model: Its past and its Journal of Management, 22(4), 542–555.
future in health care. Journal of Biomedical Informatics, 43(1), 159–172. https://doi. Pituch, K. A., & Lee, Y. K. (2006). The influence of system characteristics on e-learning
org/10.1016/j.jbi.2009.07.002 use. Computers & Education, 47(2), 222–244.
Hoque, M. R., Saif, A. N. M., AlBar, A. M., & Bao, Y. (2016). Adoption of information and Ragasa, C., Ulimwengu, J., Randriamamonjy, J., & Badibanga, T. (2015). Factors
communication technology for development: A case study of small and medium affecting performance of agricultural extension: Evidence from Democratic Republic
enterprises in Bangladesh. Information Development, 32(4), 986–1000. https://doi. of Congo. The Journal of Agricultural Education and Extension, 22(2), 113–143.
org/10.1177/0266666915578202 https://doi.org/10.1080/1389224X.2015.1026363
Igbaria, M., Parasuraman, S., & Baroudi, J. J. (1996). A motivational model of Robinson, J. P., Shaver, P. R., & Wrightsman, L. S. (1991). Criteria for scale selection and
microcomputer usage. Journal of Management Information Systems, 13(1), 127–143. evaluation. In J. P. Robinson, P. R. Shanver, & L. S. Wrightsman (Eds.), Measures of
Islam, M. S., Haque, M. E., Afrad, M. S. I., Abdullah, H. M., & Hoque, M. Z. (2017). personality and social psychological attitudes (pp. 1–13). San Diego, CA: Academic
Utilization of ICTs in agricultural extension services of Bangladesh. Asian Journal of Press.
Agricultural Extension, Economics & Sociology, 16(1), 1–11. Sakil, A. H. (2018). ICT, youth and urban governance in developing countries:
Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39(1), 31–36. Bangladesh perspective. International Journal of Adolescence and Youth, 23(2),
Kamruzzaman, M., Chowdhury, A., Paassen, A. V., & Ganpat, W. (2018). Extension 219–234. https://doi.org/10.1080/02673843.2017.1330697
agents’ use and acceptance of social media: The case of the department of Sammah, B. A., Azril, H., Shaffril, M., Hassan, M. A., & D’Silva, J. L. (2011). Can
agricultural extension in Bangladesh. Journal of International Agricultural and technology acceptance model be applied on the rural setting: The case of village
Extension Education, 25(2), 132–149. https://doi.org/10.5191/jiaee.2018.25210 development and security committee in Malaysia. Journal of Social Sciences, 7(2),
Kamruzzaman, M., Chowdhury, A., Odame, H. H., & Sarapura, S. (2019). How do 113–119. https://doi.org/10.3844/jssp.2011.113.119
extension agents of DAE use social media for strengthening agricultural innovation Sein, M. K., Thapa, D., Hatakka, M., & Sæbø, Ø. (2019). A holistic perspective on the
in Bangladesh? Rural Extension and Innovation Systems Journal, 15(1), 10–19. theoretical foundations for ICT4D research. Information Technology for Development,
Kantamara, P., & Vathanophas, V. (2014). Single-loop vs. double-loop learning: An 25(1), 7–25. https://doi.org/10.1080/02681102.2018.1503589
obstacle or a success factor for organisational learning. International Journal of Sivakumar, P. S., Parasar, B., Das, R. N., & Anantharaman, M. (2014). Determinants of
Educational Research, 2(7), 55–62. computer utilization by extension personnel: A structural equations approach. The
Kim, H., Chan, H. C., & Gupta, S. (2007). Value-based adoption of mobile internet: An Journal of Agricultural Education and Extension, 20(2), 191–212. https://doi.org/
empirical investigation. Decision Support Systems, 43(1), 111–126. https://doi.org/ 10.1080/1389224X.2013.803986
10.1016/j.dss.2005.05.009 Smith, M. R., & Kotrlik, J. W. (1990). Computer anxiety levels of southern region
Klerkx, L., Hall, A., & Leeuwis, C. (2009). Strengthening agricultural innovation capacity: cooperative extension agents. Journal of Agricultural Education, 31(1), 12–17.
Are innovation brokers the answer? International Journal of Agricultural Resources, Steinke, J., Van Etten, J., Müller, A., Ortiz-crespo, B., De Gevel, V., Silvestri, S., &
Governance and Ecology, 8(5–6), 409–438. https://doi.org/10.1504/ Priebe, J. (2020). Tapping the full potential of the digital revolution for agricultural
IJARGE.2009.032643 extension: An emerging innovation agenda. International Journal of Agricultural
Kline, R. B. (2016). Principles and practice of structural equation modeling (4th ed.). New Sustainability, 0(0), 1–17. https://doi.org/10.1080/14735903.2020.1738754
York: Guilford Press. Strong, R., Ganpat, W., Harder, A., Irby, T. L., & Lindner, J. R. (2014). Exploring the use
Lee, Y.-H., Hsieh, Y.-C., & Chen, Y.-H. (2013). An investigation of employees’ use of e- of information communication technologies by selected Caribbean extension
learning systems: Applying the technology acceptance model. Behaviour & officers. The Journal of Agricultural Education and Extension, 20(5), 485–495. https://
Information Technology, 32(2), 173–189. https://doi.org/10.1080/ doi.org/10.1080/1389224X.2014.927373
0144929X.2011.577190 Suvedi, M., Ghimire, R., & Kaplowitz, M. (2017). Farmers’ participation in extension
Legris, P., Ingham, J., & Collerette, P. (2003). Why do people use information programs and technology adoption in rural Nepal: A logistic regression analysis. The
technology? A critical review of the technology acceptance model. Information & Journal of Agricultural Education and Extension, 23(4), 351–371. https://doi.org/
Management, 40(3), 191–204. https://doi.org/10.1016/S0378-7206(01)00143-4 10.1080/1389224X.2017.1323653
Maredia, M. K., Reyes, B., Ba, M. N., Dabire, C. L., Pittendrigh, B., & Bello-Bravo, J. Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston, MA:
(2018). Can mobile phone-based animated videos induce learning and technology Pearson/Allyn & Bacon.
adoption among low-literate farmers? A field experiment in Burkina Faso. Taherdoost, H. (2018). Development of an adoption model to assess user acceptance of e-
Information Technology for Development, 24(3), 429–460. https://doi.org/10.1080/ service technology: E-service technology acceptance model. Behaviour & Information
02681102.2017.1312245 Technology, 37(2), 173–197. https://doi.org/10.1080/0144929X.2018.1427793
Maria, K., Maria, B., & Andrea, K. (2021). Exploring actors, their constellations, and roles Tata, J. S., & McNamara, P. E. (2016). Social factors that influence use of ICT in
in digital agricultural innovations. Agricultural Systems, 186, Article 102952. https:// agricultural extension in southern Africa. Agriculture, 6(2), 15. https://doi.org/
doi.org/10.1016/j.agsy.2020.102952 10.3390/agriculture6020015
Materia, V. C., Giarè, F., & Klerkx, L. (2015). Increasing knowledge flows between the Tata, J. S., & McNamara, P. E. (2018). Impact of ICT on agricultural extension services
agricultural research and advisory system in Italy: Combining virtual and non-virtual delivery: Evidence from the catholic relief services SMART skills and farmbook
interaction in communities of practice. The Journal of Agricultural Education and project in Kenya. The Journal of Agricultural Education and Extension, 24(1), 89–110.
Extension, 21(3), 203–218. https://doi.org/10.1080/1389224X.2014.928226 https://doi.org/10.1080/1389224X.2017.1387160
Miller, C., Saroja, V. N., & Linder, C. (2013). ICT uses for inclusive agricultural value chains. Todd, P. R., & Javalgi, R. R. G. (2007). Internationalization of SMEs in India: Fostering
Rome: FAO. Retrieved from http://www.fao.org/3/a-aq078e.pdf. entrepreneurship by lever- aging information technology. International Journal of
Molla.. (2020). Tech companies tried to help us spend less on our phones. It didn't work. Emerging Markets, 2(2), 166–180.
source https://www.vox.com/recode/2020/1/6/21048116/tech-companies-time-w Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology
ell-spent-mobile-phone-usage-data. acceptance model: Four longitudinal field studies. Management Science, 46(2),
Mugwisi, T., Mostert, J., & Ocholla, D. N. (2015). Access to and utilization of information 186–204.
and communication technologies by agricultural researchers and extension workers Verma, P., & Sinha, N. (2018). Integrating perceived economic wellbeing to technology
in Zimbabwe. Information Technology for Development, 21(1), 67–84. https://doi.org/ acceptance model: The case of mobile based agricultural extension service.
10.1080/02681102.2013.874317 Technological Forecasting and Social Change, 126, 207–216. https://doi.org/10.1016/
Mukta, M. Z. N., Karim, A. S. M. Z., & Kashem, M. A. (2010). Cell phone use by the j.techfore.2017.08.013
farmers in securing need-based information. Bangladesh Journal of Extension Yi, M. Y., & Hwang, Y. (2003). Predicting the use of web-based information systems: Self-
Education, 22(1&2), 7–14. efficacy, enjoyment, learning goal orientation, and the technology acceptance
Munoz-Leiva, F., Climent-Climent, S., & Leibana-Cabanillas, F. (2017). Determinants of
intention to use the mobile banking apps: An extension of the classic TAM model.
Spanish Journal of Marketing - ESIC, 21(1), 25–38. https://doi.org/10.1016/j.
sjme.2016.12.001

12
K.H. Kabir et al. Digital Geography and Society 3 (2022) 100027

model. International Journal of Human-Computer Studies, 59(4), 431–449. https://doi. Zijlmans, E. A. O., Tijmstra, J., Van der Ark, L. A., & Sijtsma, K. (2018). Item-score
org/10.1016/S1071-5819(03)00114-9 reliability in empirical-data sets and its relationship with other item indices.
Zhang, J., Zhang, X., Mu, W., Zhang, J., & Fu, Z. (2009). Farmers information usage Educational and Psychological Measurement, 78(6), 998–1020. https://doi.org/
intention in China based on the technology acceptance model. Computer and 10.1177/0013164417728358
Computing Technologies in Agriculture II, 3, 1845–1853.

13